neuroscience and biobehavioral reviewsbrainmap.org/pubs/philip_nbr_12.pdf · 2015. 1. 6. · r.c.m....

R

As

RSa

b

a

ARRA

KAFSM

C

T

0d

Neuroscience and Biobehavioral Reviews 36 (2012) 901–942

Contents lists available at SciVerse ScienceDirect

Neuroscience and Biobehavioral Reviews

jou rna l h omepa ge: www.elsev ier .com/ locate /neubiorev

eview

systematic review and meta-analysis of the fMRI investigation of autismpectrum disorders

uth C.M. Philipa,1, Maria R. Dauvermanna,1, Heather C. Whalleya, Katie Baynhama,tephen M. Lawriea, Andrew C. Stanfielda,b,∗

Division of Psychiatry, University of Edinburgh, Kennedy Tower, Royal Edinburgh Hospital, Edinburgh, EH10 5HF, UKPatrick Wild Centre, University of Edinburgh, Kennedy Tower, Royal Edinburgh Hospital, Edinburgh, EH10 5HF, UK

r t i c l e i n f o

rticle history:eceived 2 July 2010eceived in revised form 24 October 2011ccepted 26 October 2011

a b s t r a c t

Recent years have seen a rapid increase in the investigation of autism spectrum disorders (ASD) throughthe use of functional magnetic resonance imaging (fMRI). We carried out a systematic review and ALEmeta-analysis of fMRI studies of ASD. A disturbance to the function of social brain regions is among the

eywords:utism spectrum disordersunctional magnetic resonance imaging

most well replicated finding. Differences in social brain activation may relate to a lack of preference forsocial stimuli as opposed to a primary dysfunction of these regions. Increasing evidence points towardsa lack of effective integration of distributed functional brain regions and disruptions in the subtle modu-lation of brain function in relation to changing task demands in ASD. Limitations of the literature to dateinclude the use of small sample sizes and the restriction of investigation to primarily high functioning
ystematic review
eta-analysis males with autism.© 2011 Elsevier Ltd. All rights reserved.

ontents

1. Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9022. Methods . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 902

2.1. Literature search methods . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9022.2. Inclusion criteria . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9022.3. Data extraction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9022.4. ALE meta-analysis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 902

3. Results . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9033.1. Included papers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9033.2. Details of ASD participants . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9033.3. Behavioural performance . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9033.4. ALE meta-analyses . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9033.5. Motor tasks . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9033.6. Visual processing tasks . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9033.7. Executive function tasks . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9033.8. Auditory and language tasks . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9053.9. Basic social processing tasks . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9053.10. Complex social cognition tasks . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9063.11. Connectivity . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 906

4. Discussion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9094.1. Methodology of included literature . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 909

4.1.1. Study participants . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

4.1.2. Image analysis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

4.1.3. In-scanner behavioural performance . . . . . . . . . . . . . . . . . . . . .

∗ Corresponding author at: Division of Psychiatry, University of Edinburgh, Kennedy Toel.: +44 1315376265; fax: +44 1315376531.

E-mail address: [email protected] (A.C. Stanfield).1 These authors contributed equally to the paper.

149-7634/$ – see front matter © 2011 Elsevier Ltd. All rights reserved.oi:10.1016/j.neubiorev.2011.10.008

. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 909

. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 913 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 917

wer, Royal Edinburgh Hospital, Edinburgh, EH10 5HF, UK.

dx.doi.org/10.1016/j.neubiorev.2011.10.008

http://www.sciencedirect.com/science/journal/01497634

http://www.elsevier.com/locate/neubiorev

mailto:[email protected]

dx.doi.org/10.1016/j.neubiorev.2011.10.008

9

1

crPittmdrucmwgnmaacs

pAsA

2

2

l2sdcosuo

02 R.C.M. Philip et al. / Neuroscience and Biobehavioral Reviews 36 (2012) 901–942

4.2. Task related activation differences . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9174.2.1. Motor tasks . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9174.2.2. Visual processing tasks . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9254.2.3. Executive function tasks . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9304.2.4. Auditory and language tasks. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9354.2.5. Basic social processing . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9354.2.6. Complex social cognition tasks . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 936

4.3. Emerging themes across task domains . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9374.3.1. The influence of specific task/stimuli demands . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9374.3.2. Lack of preference for social stimuli . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9374.3.3. Lack of modulation in response to task/stimuli demands . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9374.3.4. Visuo-spatial processing style . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9384.3.5. Dysconnectivity . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9384.3.6. Age-related changes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 938

4.4. Limitations of the current review . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9385. Future directions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9386. Conclusion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 939

Acknowledgments. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 939References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 939

. Introduction

Autism spectrum disorders (ASD) are characterised by diffi-ulties in social interactions, communication impairments and aestricted repetitive pattern of interests and behaviours (Americansychiatric Association, 2000). Neuropsychological studies havedentified many differences between individuals with ASD and con-rols, in particular differences in sensory processing, disturbanceso executive function and a reduced capacity to appreciate the

ental states of others (Rajendran and Mitchell, 2007). Whilst theifficulties which are prevalent in ASD are largely accepted as neu-odevelopmental in origin, the changes to brain function whichnderlie the conditions are only recently being recognised. A majorontributor to this increase in understanding has been the develop-ent of functional magnetic resonance imaging (fMRI), a techniquehich exploits differences in the ferromagnetic properties of oxy-

enated and deoxygenated blood to produce an indirect measure ofeuronal activity (Cohen and Bookheimer, 1994). Since its develop-ent, fMRI has been widely applied in individuals with ASD using

variety of populations, tasks and methods of analysis. This reviewnd meta-analysis aims to draw together this literature in order toharacterise the populations examined and the methodology usedo that common findings from across studies can be identified.

The aim of this review is to (1) assess the ASD populations takingart in fMRI research in terms of how representative they are of theSD population as a whole; (2) identify common findings acrosstudies which indicate brain regions which function differently inSD, relative to typically developed individuals.

. Methods

.1. Literature search methods

Medline, EMBASE and PsychINFO were searched for all Englishanguage studies published between January 1984 and August009 that reported functional MRI data in people with an autismpectrum disorder. Search terms included ‘autism’, ‘Asperger Syn-rome’, ‘pervasive developmental disorder’ and related terms wereombined using the AND operator with ‘functional magnetic res-

2.2. Inclusion criteria

Articles were included if they were primary research studiespublished as peer-reviewed articles in English and they compareda sample of participants with an autism spectrum disorder with agroup of neurotypical controls, using fMRI. Abstracts were assessedfor inclusion, and full text articles were retrieved where appropri-ate.

2.3. Data extraction

For each study, data for the ASD participant group wereextracted: gender, mean age and mean IQ. Where possible, datapertaining to the diagnosis of the ASD group was also extracted,including the diagnostic criteria used. The selection of the compar-ison group and features by which they were matched to the ASDgroup was also recorded. Details of the fMRI paradigm were alsoextracted and studies grouped according to the element of cogni-tion under investigation. The fMRI studies were allocated to six taskdomains: motor tasks; visual processing tasks; executive functiontasks; auditory and language tasks; basic social processing tasks(face processing, emotion processing, motion in relation to socialstimuli, eye gaze) and complex social cognition tasks (imitation,irony comprehension, empathy). Many tasks contain aspects frommultiple domains and where this occurred the tasks were con-sidered in more than one analysis. fMRI studies, which analysedfunctional and effective connectivity, were separately considered.

Within scanner performance was examined by extracting accu-racy and reaction time data where available for the ASD and thecontrol groups. For the meta-analysis, coordinates where the BOLDresponse differed significantly between the ASD and control groupswere extracted where available.

2.4. ALE meta-analysis

Statistically significant foci from between group contrasts wererecorded for each study. MNI coordinates were converted toTalairach coordinates. The meta-analyses were conducted withactivation likelihood estimation (ALE 2.0). The ALE algorithm isbased on models (Eickhoff et al., 2009) implemented in BrainMap

nance imaging’ OR ‘fMRI’. Both free-text and expanded medicalubject headings were used. The search strategy was supplementedsing a cited reference search and by inspecting the reference listsf included articles.

(Laird et al., 2005). ALE models the activation foci as a three-dimensional Gaussian probability density function centred at thegiven coordinates. In the next step ALE calculates the spatial over-lap of these distributions across different experiments. The spatial

Biobehavioral Reviews 36 (2012) 901–942 903

uiaeme

tswtc1ml

3

3

icfs2imTpa

3

twAmstraoccstai

dbdaPwdt(SaaIa

Table 1Number of experiments and foci included in the analysis of each task domain.

Group Number ofexperiments

Total numberof subjects

Numberof foci

Motor tasks 3 24 107Visual processing tasks 4 40 83Executive processing tasks 10 116 118Auditory and language tasks 10 125 181

R.C.M. Philip et al. / Neuroscience and

ncertainty associated with activation foci is estimated to thenter-subject and inter-experiment variability. The convergence ofctivation patterns across studies is computed by taking the mod-lled activation maps. The estimation was constrained by a greyatter mask and estimated the above chance clustering with the

xperiments as a random-effects factor (Eickhoff et al., 2009).For verification of ASD-related differences in activation for each

ask domain, we performed ALE analyses (ASD > C, C > ASD) for allix task domains separately. In addition, in the three task domainshere studies with children and adolescents were available (audi-

ory and language tasks, basic social tasks and complex socialognition tasks), separate analyses of children/adolescents (under8 years old) and adults (over 18 years old) were carried out. Resultap overlays were produced on a standardised structural scan for

ocalisation.

. Results

.1. Included papers

95 Papers were identified which reported using fMRI tonvestigate ASD. One article performed spectroscopy analysis inonjunction with fMRI data presented elsewhere and was there-ore excluded (Kleinhans et al., 2007). Three papers presented casetudies so were also excluded (Turkeltaub et al., 2004; Grelotti et al.,005; Carmody et al., 2007). Finally, one study was excluded as it

nvestigated a group of participants with autism pre and post treat-ent and did not investigate a comparison group (Bölte et al., 2006).

his left 90 original research papers in which a group of partici-ants with ASD were investigated using functional MRI paradigmsnd compared to a control group.

.2. Details of ASD participants

The largest sample size was a pooled analysis of data collectedogether from several other studies and concerned 57 participantsith ASD. The largest original study contained 19 participants withSD, whilst the smallest concerned only 5 individuals. Overall theean number of participants with ASD per study was 12. Seven

tudies failed to report the gender ratio in their sample. Of thosehat did, 60% of studies investigated only male participants. Theatio of males to females participating across all fMRI studies ispproximately 15:1. One study did not provide details of the agef the participants in their study. Of the remaining 89 studies, 24oncerned participants with a mean age of less than 18 and 65 con-erned participants with a mean age of greater than 18. Almost alltudies provided intelligence quotient (IQ) scores for the ASD par-icipants in their study. The median full scale IQ score was 104, with

reported range of 55–139. The majority of studies however onlyncluded participants with an IQ over 70.

In total, 1083 participants with an autism spectrum disor-er were reported on, although there was considerable overlapetween studies. In 13 studies, a total of 138 participants wereescribed as an unspecified combination of individuals withutism/high functioning autism and/or Asperger Syndrome and/orDD-NOS. In a further 10 studies, a total of 133 participants,ere described as having an ASD, presumably encompassing theisorders mentioned above. The remainder of studies specifiedhe diagnosis of participants; 609 had a diagnosis of autismmainly ‘high-functioning autism’), 188 had a diagnosis of Aspergeryndrome and 15 had a diagnosis of PDD-NOS. Expert clinical
ssessment was used for participant diagnosis which was gener-lly supported by the use of DSM-IV criteria (in 51 studies) andCD-10 criteria (a further 9 studies). The ADI (Lord et al., 1994)nd ADOS (Lord et al., 1989) were commonly used as standardised
Basic social tasks 20 253 244Complex social cognition tasks 9 122 118

diagnostic measures, in 73 and 65 studies respectively. However,according to some of the published scores, participants did notalways meet the cut-off criteria on these instruments or scoresfor some participants were missing. In studies with younger par-ticipants the Childhood Autism Rating Scale assessment (CARS;Schopler et al., 1980) and the Vineland Adaptive Behaviour Scale(Sparrow and Cicchetti, 1985) were applied. The Autism SpectrumQuotient (AQ; Woodbury-Smith et al., 2005) and Social Respon-siveness Scale (SRS; Constantino et al., 2003) were also used ascontinuous measures of autistic traits in three studies.

3.3. Behavioural performance

No group differences in accuracy and/or reaction time betweenthe ASD and the controls were found in most of the task domainsfor children/adolescents and adults. There was a group differencein accuracy and reaction time in complex social cognition tasks inadults, (F(1,28) = 7.40, p < 0.05).

3.4. ALE meta-analyses

49 Studies provided co-ordinates where significant differencesbetween individuals with ASD and controls were identified. In themain analysis, across the six task domains there were 851 foci from85 contrasts (see Table 1). In the three domains where separateage-related analyses were carried out 235 foci from 24 experi-ments were included for children/adolescents and 253 foci from29 experiments for adults (Tables 2 and 3, Fig. 1).

3.5. Motor tasks

Individuals with autism activated the bilateral precentral gyriand the inferior/middle frontal gyri more than the control subjects.In contrast, controls displayed greater activations of the left culmenand the right superior temporal gyrus than are evident in those withASD.

3.6. Visual processing tasks

The ASD group showed greater activation than controls in thethalamus and the medial frontal gyrus, whilst controls showedmore activation than individuals with ASD in the cingulate gyrusand occipital regions (Tables 4 and 5, Fig. 2).

3.7. Executive function tasks

Individuals with ASD showed greater activation in the left mid-dle frontal gyrus (BA11). In comparison, control subjects showed
greater activation than those with ASD in the right middle frontalgyrus (BAs 6 and 9) as well as other prefrontal regions and subcor-tical regions (Tables 6 and 7, Fig. 3).


Fig. 1. Motor tasks. ALE maps (pFDR < 0.05) are superimposed on slices from grey matter template in Talairach space. The top panel illustrates areas of greater probability ofactivation in ASD subjects compared to controls in a cluster centred at (a) left middle frontal gyrus (x = −44, y = 14, z = 44) and right inferior forntal gyrus (x = 48, y = 10, z = 22);(b) the bilateral precentral gyrus (left, x = −40, y = −6, z = 42); (right, x = 44, y = −6,42); (c) the left superior parietal lobe (x = −30, y = −70, z = 50). The bottom panel showsC > ASD activation likelihood estimate maps in clusters centred at (d) right lentiform nucleus (x = 26, y = 4, z = −4) and right middle frontal gyrus (x = 30, y = 6, z = 52); (e) leftprecentral gyrus (x = −36, y = −22, z = 54); (f) right superior temporal gyrus (x = 64, y = −44, z = 18), (g) the left culmen (x = −12, y = −58, z = −6) and right inferior parietal lobe(x = 40, y = −54, z = 48).

Fig. 2. Visual processing tasks. ALE maps (pFDR < 0.05) are superimposed on slices from grey matter template in Talairach space. The top panel illustrates areas of greaterprobability of activation in ASD subjects compared to controls in a cluster centred at (a) left thalamus (x = 0, y = −16, z = 4); (b) left medial frontal gyrus (x = −8, y = −12, z = 58)and left caudate (x = −4, y = 10, z = 6). The bottom panel shows C > ASD activation likelihood estimate maps in clusters centred at (c) left cingulate gyrus (x = 0, y = 10, z = 40);(d) left precentral gyrus (x = −36, y = −14, z = 50), (e) left middle occipital gyrus (x = −48, y = −66, z = −8); left occipital lingual gyrus (x = 0, y = −78, z = 4).

Fig. 3. Executive function tasks. ALE maps (pFDR < 0.05) are superimposed on slices from grey matter template in Talairach space. The top panel illustrates areas of greaterprobability of activation in ASD subjects compared to controls in a cluster centred at (a) left middle frontal gyrus (x = −28, y = 42, z = −8). The bottom panel shows C > ASDactivation likelihood estimate maps in clusters centred at (b) right middle frontal gyrus (x = 40, y = 4, z = 50); (c) left lentiform nucleus (x = −20, y = 10, z = 2); (d) right middlefrontal gyrus (x = 48, y = 20, z = 28); (e) left insula (x = −40, y = −12, z = 12); (f) left inferior parietal lobule (x = −34, y = −48, z = 46) and right posterior cingulate (x = 14, y = −48,z = 8).

R.C.M. Philip et al. / Neuroscience and Biobehavioral Reviews 36 (2012) 901–942 905

Table 2Motor tasks. The main findings reported are significant differences in a whole brain between group contrast unless detailed in italics or otherwise stated. Abbreviations; ADI-R– autism diagnostic interview-revised, ADOS – autism diagnostic observational schedule, AQ – autism spectrum quotient, AS – Asperger symdrome, DSM-IV – diagnostic andstatistical manual of mental disorders – 4th edition, HFA – high functioning autism, ICD-10 – international classification of diseases – 10th edition, IQ – intelligence quotient,PDD-NOS – pervasive developmental disorder – not otherwise specified, ROI – region of interest.

Study (year) Autism group Control matchingcriteria

Task design Main findings

N (M:F) Mean age(sd)

Mean IQ(sd)

Diagnosis Diagnosticmeasures

Allen andCourchesne(2003)

7:1 26.9 (8.6) 85 (11.95) AutismDSM-IV, ADI-R,ADOS

Age, gender,handedness, taskperformance

1. Self paced buttonpress; ‘rest’ as baseline.

1. ASD > control:cerebellum.

2. Button press totarget; passive viewingof visual stimuli asbaseline.

2. Control > ASD:cerebellum.(ROI analysis)

Allen et al.(2004)

7:1 26.9 (8.6) 85 (11.95) Autism DSM-IV, ADI-R,ADOS


Self paced buttonpress; ‘rest’ as baseline.

ASD > control:ipsilateral anteriorcerebellum.(ROI analysis)

Muller et al.(2001)*

8:0 28.4 (8.9) 86.5 (11.4) Autism DSM-IV, ADI-R,CARS


Visually paced motorresponse; passiveviewing of visualstimuli.

Control > ASD:contralateralsensorimotor cortexand anterior temporallobe including anteriorinsula and caudate.ASD > control: bilateralparieto-occipital andcontralateral prefrontalcortex.

Muller et al.(2003)* 8:0 28.4 (8.9) 86.5 (11.4) Autism

DSM-IV, ADI-R,CARS

Age, gender,handedness,

1. Image of hand withdot indicatingappropriate buttonpress for a 6-digitrepeated sequence;single-digit stimuli asbaseline.

1. Control > ASD:bilateral occipital andsuperior parietal cortexand right middlefrontal gyrus.ASD > control: bilateralparietal lobes,premotor cortex, rightmedial frontal area andleft middle andsuperior frontal gyri.

2. Image of hand withdot indicatingappropriate buttonpress for a 6-digitrepeated sequence;regular 6-digitsequence as baseline.

2. Control > ASD:bilateral premotor,superior parietalanterior inferiorparietal,tempero-occipitalcortex and left anteriorcerebellum.ASD > control: frontalcortex anterior topremotor cortex andinferior and posteriorparietal cortex.

Muller et al.(2004)*

8:0 28.4 (8.9) 86.5 (11.4) Autism DSM-IV, ADI-R,CARS

Age, gender,handedness

Image of hand with dotindicating appropriatebutton press for a8-digit repeatedsequence; single-digit

‘late learning’ vs ‘earlylearning’Control > ASD;prefrontal cortexASD > control; right

3

gpci

cgws

* Study included in meta-analysis.

.8. Auditory and language tasks

When all auditory and language studies were included, the ASDroup showed greater likelihood of activated clusters in the rightrecentral gyrus and the left declive compared to the controls. Inontrast, controls showed greater activation than those with ASDn the bilateral superior temporal gyri (Tables 8 and 9, Fig. 4).

When separated by age, relative overactivation in the right pre-
entral gyrus was seen in children/adolescents with ASD, whilstreater activation in the bilateral declive was apparent in adultsith ASD. Unaffected children/adolescents activated the bilateral
uperior temporal gyri significantly more than those with ASD. In

stimuli as baseline. pericentral andpremotor cortex

both age groups, control subjects displayed greater probability ofactivation of the left cingulate gyrus. Adults with ASD showed moreclusters of relative overactivation than underactivation whilst thereverse was true in children/adolescents with ASD.

3.9. Basic social processing tasks

Within the whole group analysis for basic social processingtasks the individuals with ASD showed greater likelihood than con-trols of activated clusters in the bilateral superior temporal gyri,whilst controls showed greater activation than those with ASD in


Table 3Group comparisons of regions with significantly elevated likelihood of activation in motor tasks. Brain areas activated from the ALE analysis (pFDR < 0.05 and a minimumcluster size of 200 voxels).

Comparison – age group Brain region BA Volume (mm3) Talairach ALE (10−2)

x y z

ASD > C; Adults

Superior Parietal Lobule L 7 944 −30 −70 50 0.90Precentral Gyrus R 6 752 44 −6 42 1.03Middle Frontal Gyrus L 8 488 −44 14 44 1.03Inferior Frontal Gyrus R 9 432 48 10 22 1.02Precentral Gyrus L 6 264 −40 −6 42 0.80

C > ASD; Adults

Culmen L 912 −12 −58 −6 1.24Inferior Parietal Lobule R 40 560 40 −54 48 1.08Superior Temporal Gyrus R 22 328 64 −44 18 0.97

cg

cstgfb

w

3

awttcF

tpvfe

Fiplg

Lentiform Nucleus RMiddle Frontal Gyrus R 6

Precentral Gyrus L 4

lusters in the left fusiform gyrus and the right inferior occipitalyrus (Tables 10 and 11, Fig. 5).

Adults with ASD showed a higher likelihood of activation thanontrols in a cluster representing social regions like the bilateraluperior temporal gyri. Controls had a greater probability of activa-ion in the right parahippocampal gyrus and right inferior occipitalyrus, thought to be object-related processing regions, and the leftusiform gyrus. Regarding the distribution and numbers of clustersetween the age groups, adults with ASD showed a mixture of over-

and under-activation compared to controls whereas childrenith ASD tended to underactivate.

.10. Complex social cognition tasks

When all complex social cognition tasks were entered into thenalysis, the left superior temporal gyrus appeared as a region thatas both over- and underactivated in individuals with ASD relative

o controls. In addition a significant cluster in the right superioremporal gyrus was seen as a region of greater activation in theontrol group compared to individuals with ASD (Tables 12 and 13,ig. 6).

Children and adolescents with ASD showed greater activa-ion than controls in the left inferior frontal gyrus, left pre- and
ost-central gyri and the left superior temporal gyrus. Reduced acti-ation in children/adolescents with ASD was seen in left superiorrontal gyrus, right superior temporal gyrus and left inferior pari-tal lobule, whereas adults with ASD showed reduced activation
ig. 4. Auditory and language tasks. Whole group analysis. ALE maps (pFDR < 0.05) are sullustrates areas of greater probability of activation in ASD subjects compared to controrecentral gyrus (x = 36, y = −10, z = 54); (c) left posterior cingulate 9x = 0, y = −50, z = 14); (

ikelihood estimate maps in clusters centred at (e) left cingulate gyrus (x = 14, y = 18, z = 24yrus (x = −52, y = −20, z = 10); (h) right pyramis (x = 10, y = −78, z = −24).

288 26 4 −4 0.96248 30 6 52 0.85216 −36 −22 54 0.85

in the right claustrum. The age-related pattern of significant clus-ters in this task domain is different than in the other task domainswith children and adolescents with ASD showing more areas ofboth over- and under-activation relative to controls than are seenin adults with ASD

3.11. Connectivity

ALE meta-analysis is not possible for studies of functional andeffective connectivity. However, the systematic review showedthat all but two of the studies (Welchew et al., 2005; Noonan et al.,2009) examining brain connectivity identified reductions in con-nectivity in individuals with ASD compared to controls. Of these, themajority identified underconnectivity alone although three studies(Mizuno et al., 2006; Turner et al., 2006; Wicker et al., 2008) identi-fied a mixed pattern of increases and decreases. Both Mizuno et al.(2006) and Turner et al. (2006) found atypically increased activa-tions in subcortical regions in ASD whereas Wicker et al. reporton altered cortico-cortical connectivity. Reductions in connectiv-ity in individuals with ASD were identified across a wide varietyof brain areas using tasks which probed aspects of executive func-tion (Koshino et al., 2005, 2008; Just et al., 2006; Kana et al., 2007;
Kleinhans et al., 2008b), language (Just et al., 2004; Mason et al.,2008), emotion processing (Wicker et al., 2008) and motor func-tions (Villalobos et al., 2005; Mizuno et al., 2006; Turner et al.,2006). In addition, two studies identified reductions in resting state
perimposed on slices from grey matter template in Talairach space. The top panells in a cluster centred at (a) left frontal sub-gyral (x = −20, y = 2, z = 52); (b) rightd) left declive (x = −18, y = −78, z = −18). The bottom panel shows C > ASD activation); (f) right superior temporal gyrus (x = 56, y = −8, z = −2); (g) left superior temporal


Table 4Visual processing tasks. Abbreviations as in Table 2.

Study (year) Autism group Controlmatchingcriteria



Mean IQ(sd)


Bölte et al.(2008) 7:0 27.7 (7.8) 98 (19.2) Autism

ICD-10, ADI-R,ADOS

Age, IQ, gender,handedness,taskperformance

1. Count triangleswithin Block Designstimuli; fixationbaseline.

1. Control > ASD: rightventral quadrant ofprestriate visual cortex(V2v) and ventralposterior visual cortex(VP).

2. Counting colours;fixation baseline.

2. Control > ASD:ventral posterior visualcortex (VP).(ROI analysis)

Hadjikhaniet al. (2004b)

8 35 (12) 117 (6) Autism, AS,PDD-NOS

ADI-R, ADOS IQ Visual checkerboard No qualitativedifference in activationpatterns betweengroups.

Hubl et al.(2003)

7:0 27.7 (7.8) 98 (17) Autism ICD-10, ADI,ADOS

Age, IQ, gender,taskperformanceaccuracy (notRT)

1. Indentify target viabutton press. Happy,sad, angry and neutralfaces presented. Targethappy (explicitemotion condition);target female (implicitemotion condition);scrambled faces asbaseline condition.2. Colour counting;shape counting withina mosaic; and a restcondition.

Values from ROIs wereextracted from eachcondition andinvestigated in anANOVA for interactionswith diagnosis, task,region and hemisphere.Activations for each taskwere different betweenASD and controls.Different tasks activateddifferent regions.The ASD and controlgroups differ withrespect to the differencein activations caused byeach task in differentregions.

Keehn et al.(2008)*

9:0 15.1 (2.6) PIQ 110 (20) ASD DSM-IV, ADI,ADOS (but onedid not meetcut off)

Age, gender,PIQ

Visual search paradigmwhere participantsindicated via buttonpress whether thetarget was present ornot.There were 12 trialtypes;presence/absence oftarget, homogen-sous/heterogenousdistractors, 3 set sizes.High level baseline;solitary target ordistractor.Low level baseline;visual fixation.

High level baseline vslow level baselineASD > control: rightinferior frontal gyrusHomogenous/heterogenoustrials vs high levelbaselineASD > controls: rightmiddle occipital gyrus,inferior frontal gyrusand middle occipitalgyrus. Leftsupplementary motorarea, superior parietallobe and precentralgyrus.

Lee et al.(2007b)

12:5 10.37 (1.52) 109.3 (14.2) 8 HFA, 9 AS DSM-IV, ADI-R,ADOS

Age, IQ, taskperformance

Embedded figures task;shape matching task asbaseline.

No significant groupdifferences.Qualitative difference inactivation mapsbetween groups; ASDgroup failed to recruitmedial and lateralprefrontal cortex,ventral temporal cortexand inferior parietalcortex; parietal andoccipital activation wasbilateral in controlgroup and unilateral inthe ASD group.

Luna et al.(2002) 9:2 32.3 (9.3) 102.7 (12.1) Autism ADI, ADOS Age, IQ

1. Visually guidedsaccades; fixationbaseline.

1. No group differences.(Groups taskperformance matched)


Table 4 (Continued)




Mean IQ(sd)


2. Ocularmotor delayedresponse task; visuallyguided saccades.

2. Control > ASD:bilateral dorsolateralprefrontal cortex andposterior cingulatecortex. (Groups nottask performancematched, ROI analysis)

Manjaly et al.(2007)

12 14.4 (2.7) 110.1 (20) 3 HFA, 9 AS DSM-IV,ICD-10, ADI-R,ADOS

Age, IQ, gender,handedness,taskperformance

Embedded figures task;visuospatial controltask as baseline.

No significantdifferences betweengroups.Qualitative difference inactivation mapsbetween groups; moreactivation in ASD groupin extrastriate cortexand calcarine sulcusthan controls.

Ring et al.(1999)*

4:2 26.3 (2.1) 108.5 (10.5) Autism, AS DSM-IV, ICD-10 Age, IQ,handedness,socioeconomicstatus,education, taskperformance

Embedded figures task;fixation baseline.

Control > ASD: bilateralparietal regions andoccipital cortex andright dorsolateralprefrontal cortex.ASD > control: rightoccipital cortexextending into inferiortemporal gyrus.

Sahyoun et al.(2009) 10:2 13.3 (2.1) 100.8 (12.3) HFA DSM-IV, ADI-R

Age, IQ,handedness,behaviouralperformance inscanner

Visuospatial andlinguistic reasoningtask. Participants filledin the blank in amatrix. Visual fixationbaseline.

1. Control > ASD: LeftMTG, lingual gyrus andsuperior precentralsulcus. Right angulargyrus.

1. visuospatial trials2. visuospatial andsemantic trials3. semantic trials

ASD > Control: Bilateralanterior MTG. Leftlateraloccipito-temporalsulcus, pre/post centralsulcus, posterior lateralfissure. Right inferiorfrontal area andpostcentral gyrus.2. Control > ASD: LeftMTG, lingual gyrus,superior precentralsulcus. Right angulargyrus, inferior frontalarea and STS.ASD > Control: Bilateralanterior MTG. Leftlateraloccipito-temporalsulcus, posterior lateralfissure, inferior parietalsulcus, occipital cortex.Right postcentralgyrus.3. Control > ASD: LeftMTG, lingual gyrus, andsuperior precentralsulcus. Rightsupermarginal gyrus,occipito-temporalcortex.ASD > Control: Leftlateraloccipito-temporalsulcus, posterior lateralfissure. Right insula,anterior MTG, inferiorfrontal area.


Table 4 (Continued)




Mean IQ(sd)


Silk et al.(2006)

7:0 14.7 (2.9) 114 (16.9) Autism, AS DSM-IV, ADI-R Age, IQ, gender,handedness

Mental rotation andmatching of shapes;shape matching asbaseline.

Control > ASD: rightinferior and medialfrontal gyri includingcaudate and dorsalpremotor cortex.

Thakkar et al.(2008)*

10:2 30 (11) 114 (9) 8 autism, 2AS, 2PDD-NOS

DSM-IV, ADI,ADOS

Age, verbal IQ,gender,handedness,socioeconomicstatus,education

Pro-saccade andanti-saccade events;fixation baseline.

1. Error vs correcteventsControl > ASD: rightmedial superior frontalgyrus.2. Correct responses vsfixationASD > control: rightrostral anteriorcingulated gyrus.

Takarae et al.(2007)* 13 24.5 (7.7) 105.9 (12.3) Autism

DSM-IV, ADI,ADOS

Age, IQ1. Visually guidedsaccades; fixationbaseline.

1. Control > ASD:bilateral frontal andsupplementary eyefields, posteriorparietal cortex andcerebellum.ASD > control: bilateraldorsolatrel prefrontalcortex, anterior andposterior cingulatecortex, medialthalamus, caudatenucleus and rightdentate nucleus.

2. Smooth pursuit of avisual target; fixationbaseline.

2. Control > ASD:bilateral frontal eyefields, posteriorparietal cortex,posterior cingulatecortex, cingulate motorarea, cerebellum,dorsolateral prefrontalcortex, precuneus andpre-supplementary

cK

4

4

4

oa

TGm


onnectivity in individuals with ASD (Cherkassky et al., 2006;ennedy and Courchesne, 2008) (Table 14).

. Discussion

.1. Methodology of included literature

.1.1. Study participantsThe majority of studies including individuals with a diagnosis

f autism used appropriate diagnostic tools to confirm this, usu-lly the ADI-R and the ADOS. It should be noted however that,

able 5roup comparisons of regions with significantly elevated likelihood of activation in visuinimum cluster size of 200 voxels).

Comparison – age group Brain region BA Vo

ASD > C; AdultsThalamus L 44Medial Frontal Gyrus L 6 36Caudate L 25

C > ASD; Adults

Cingulate Gyrus L 32 36Precentral Gyrus L 4 29Occipital Lingual Gyrus 18 26Middle Occipital Gyrus 37 22

motor area.

whilst the ADOS allows a category of autism spectrum disordersin addition to autism, the ADI-R allows only for the diagnosisof autism to be made. Confusingly several studies also reportedusing the ADI-R to confirm diagnosis in individuals with autismspectrum disorders. Also, as concepts of the autism spectrumhave broadened, increasing numbers of individuals are receiv-ing a diagnosis of PDD-NOS, which these tools are not designed
to assess. Due to differences in reporting between studies itis not possible to determine exactly the number of individu-als included in this review with PDD-NOS, but it appears to below.
al processing tasks. Brain areas activated from the ALE analysis (pFDR < 0.05 and a

lume (mm3) Talairach ALE (10−2)

x y z

8 0 −16 4 1.248 −8 −12 58 1.106 −4 10 6 0.80

8 0 10 40 1.126 −36 −14 50 0.974 0 −78 4 1.014 −48 −66 −8 0.90


Table 6Executive function tasks. Abbreviations as in Table 2.




Mean IQ (sd) Diagnosis Diagnosticmeasures

Belmonte andYurgelun-Todd(2003)*

5:1 32.7 (9.8) ‘non-retarded’ Autism, AS,PDD-NOS

DSM-IV, ADI Gender,handedness, taskperformance

Attend to onelocation and switchattention whentarget stimuliobserved; fixationbaseline.

Control > ASD:bilateral superiorparietal lobe andmedial frontal gyrus,left medial temporalgyrus, postcentralgyrus and inferiorfrontal gyrus, rightpremotor cortex andmedial frontal gyrus.

Dichter andBelger (2007) 16:1 22.9 (5.2) 105 (18.6) 14 HFA, 3 AS DSM-IV, ADI,

ADOS

Age, IQ, gender,handedness,education, taskperformance

1. Indicate viabutton pressdirection of centralarrow that wascongruent(condition1) orincongruent(condition2) withdirection of flankerarrows.

1. No significantgroup difference.

2. Indicated viabutton pressdirection of eyegaze in central facethat was congruent(condition1) orincongruent(condition2) withdirection of eyegaze in flankerfaces.

2. Control > ASD:bilateral dorsolateralprefrontal cortex,right inferiorfrontal/anteriorinsula cortex,anterior cingulateand bilateralintraparietal sulcus.(ROI analysis)

Gilbert et al.(2008) 12:3 38 (13) 119 (14) ASD ADOS


1. Sequencedbutton press;randomlygenerated buttonpress as baseline.

1. Control > ASD: leftcerebellum and leftlateral temporalcortex.

2. Press in responseto curved letters asprogress throughalphabet; sametask with only startletter presentedand thereafter taskwas stimulusindependent.

2. Control > ASD:medial parietal andoccipital cortex.ASD > control: medialprefrontal cortex,amygdala,cerebellum and othertemporal andparietal regions.

Haist et al.(2005)* 8:0 23.4 (11.4) 101.0 (9.3) 6 HFA, 2 AS DSM-IV, ADI,

ADOSAge, gender,IQ

Spatial attentiontask: two boxesand central fixationcross on the screen,initially attend tocross thenattention cued toone box followedby variable interval(short or long ISI)before letter Eappears in one ofthe boxes –congruent to cue in75% of trials,incongruent in25%; participantsasked to shiftattention to letterand note theorientation

Short ISIControl > ASD:bilateral precentraland frontal gyri, rightmiddle frontal gyrus,left parietal lobe andright insula


Table 6 (Continued)





Long ISIControl > ASD: rightmiddle frontal gyrusand left precentralgyrus

Just et al.(2007)

17:1 27.1 (11.9) 109.3 (17.7) Autism ADI, ADOS Age, IQ, gender,socioeconomicstatus, taskperformance

Indicate via buttonpress requirednumber of movesto complete Towerof London taskdisplayed, Easycondition and hardcondition; fixationbaseline.

Controls > ASD:bilateral inferior andsuperior parietalarea, angular gyri,superior, and middleoccipital areas,middle frontal gyri,and the rightprecentral gyrus,superior frontal andthe left inferiorfrontal gyri.Autism > Controls:bilateralhippocampus,thalamus and leftlingual gyrus.Autism > Controls inright middle occipitalgyrus for hard vseasy contrast.Functionalconnectivity analysisSee Table 8

Kana et al.(2007)*

11:1 26.8 (7.7) 110.1 (12.6) HFA ADI, ADOS Age, IQ, gender,handedness,socioeconomicstatus

Go/no go task withincreasing workingmemory load: 1 –simple responseinhibition, 2 –response inhibitionwith workingmemorycomponent;fixation baseline.

1. Controls > ASD: leftinferior temporalgyrus, rightparahippocampalgyrus, right calcarinesulcus, rightpremotor cortex,right middlecingulate gyrus,bilateral postcentralgyrus, rightinsula/right inferiorfrontal gyrus and leftlingual gyrus2. Controls > ASD: leftanterior cingulategyrus, left middleoccipital gyrus,bilateral calcarinesulcus, right angulargyrus and leftprecuneus3. ASD > Controls:bilateral premotorarea

Kennedy et al.(2006)*

12:0 26.5 (12.8) 101.6 (15.2) 6 autism, 6AS

ADI-R, ADOS Age Counting Strooptask withemotional, neutraland number words.Participants askedto count number ofwords on thescreen. Baselinefixation cross

Number vs restASD > Control: rightsupramarginal gyrus,right precuneus,bilateral inferiorparietal lobule, rightsuperior frontalgyrus, left medialfrontal/anteriorcingulateEmotional vs neutralControl > ASD: rightmedial orbitofrontalcortex, right middleoccipital gyrusFunctionalconnectivity analysisSee Table 8


Table 6 (Continued)





Koshino et al.(2005)*

13:1 25.7 100.1 HFA ADI, ADOS Age, IQ, gender,socioeconomicstatus, taskperformance

n-back task (0,1,2);fixation baseline.

Control > ASD: leftdorsolateralprefrontal cortex,inferior frontal gyrus,posterior precentralsulcus and inferiorparietal lobe.ASD > control: rightinferior frontal gyrus,inferior parietal lobeand bilateraltemporal lobe.(ROI analysis and nospatialnormalisation)Functionalconnectivity analysisSee Table 8


11:0 24.5 (10.2) 104.5 (13.1) HFA ADI-R, ADOS Age, IQ, gender,socioeconomicstatus, taskperformance

0, 1 and 2-backworking memorytask with faceidentity; fixationbaseline.

Control > ASD: leftinferior prefrontaland right posteriortemporal cortex.Different location ofthe Fusiform Face Areawithin the fusiformgyrus in ASD groupcompared to controls.Functionalconnectivity analysisSee Table 8

Lee et al. (2009) 9:3 10.2 (1.6) 113.3 (17.3) ASD DSM-IV,ADI-R, ADOS

Age, gender. IQ Go/NoGo task ASD > Control: rightcingulate cortexFunctionalconnectivity analysisSee Table 8

Schmitz et al.(2006)* 10:0 38 (9) 105 (14) 2 HFA, 8 AS ICD-10, ADI

Age, IQ, gender,handedness, taskperformance

1. Motor inhibition;motor response.(GO/NO-GO task)

1. ASD > control: leftmiddle/inferior andorbitofrontal gyrus.

2. Spatial STROOPtask. Incongruentevents; congruentevents.

2. ASD > control: leftinsula.

3. SWITCH task.Events where therule switched;events where therule was repeated.

3. ASD > control:right inferior and leftmesial parietalcortex.

Schmitz et al.(2008)*

10:0 37.8 (7) 107 (9) 3 HFA, 7 AS ICD-10, ADI Age, IQ, gender,handedness,socioeconomicstatus, education,task performance

Continuousperformance taskwith targetidentificationevents. Targetsassociated withmonetary reward;targets with noassociatedmonetary reward.

ASD > control: leftanterior cingulategyrus.

Shafritz et al.(2008)*

16:2 22.3 (8.7) 102.5 (17.6) Autism DSM-IV, ADI,ADOS

Age, IQ Button press inresponse totargets;non-targets asbaseline.Targets weremaintained orshifted betweenruns.

Control > ASD:dorsolateralprefrontal cortex,basal ganglia andinsula.(Stats uncorrected)



Table 7Group comparisons of regions with significantly elevated likelihood of activation in executive function tasks. Brain areas activated from the ALE analysis (pFDR < 0.05 and aminimum cluster size of 200 voxels).


x y z

ASD > C; Adults Middle Frontal Gyrus L 11 312 −28 42 −8 0.86

C > ASD; Adults Insula L 13 984 −40 −12 12 2.40Lentiform Nucleus L 912 −20 10 2 1.97Inferior Parietal Lobule L 40 704 −34 −48 46 1.39Middle Frontal Gyrus R 6 264 40 4 50 1.39

256 48 20 28 1.34224 14 −48 8 1.34

tCA2ssrmaools

iictiBdgrdlwmMe

tiaeaaaperwgiceactplc

Fig. 5. Basic social tasks. Whole group analysis. ALE maps (pFDR < 0.05) are super-imposed on slices from grey matter template in Talairach space. The top panelillustrates areas of greater probability of activation in ASD subjects compared tocontrols in a cluster centred at (a) right superior temporal gyrus (x = 54, y = 8, z = 4);(b) left superior temporal gyrus (x = −60, y = −28, z = 4). The bottom panel showsC > ASD activation likelihood estimate maps in clusters centred at (c) right culmen

Middle Frontal Gyrus R 9

Posterior Cingulate R 29

The ratio of male to female participants in the fMRI litera-ure in ASD is unrepresentative of the ASD population as a whole.ommonly reported gender ratios in autism are 4:1 or 3:1 and insperger Syndrome are 8:1 (Fombonne, 2003; Newschaffer et al.,007; Holtmann et al., 2007). The overall ratio of male to femaletudy participants in fMRI studies of 15:1 is therefore unrepre-entative of the ASD population. This is likely to be driven by theelatively small sample sizes investigated within each study whicheans researchers opt to exclude female participants to reduce

dditional variation in their sample. With the increasing numbersf participants in fMRI studies generally and perhaps also the risef multi-centre imaging (Belmonte et al., 2008), the recruitment ofarger samples that are more representative of the ASD populationhould be possible.

There is a general bias in the fMRI literature of autism towardsmaging adults. Whilst this is unsurprising given the demand-ng nature of participating in fMRI experiments, studying youngerohorts and carrying out longitudinal studies of a developmen-al disorder which manifests in early childhood is imperativef we are to gain an understanding of the aetiology of autism.oth behavioural and imaging data suggest that people with ASDevelop alternative and compensatory processing styles and strate-ies which will likely confound findings in adult populations. Theesults from our meta-analysis support the idea that results mayiffer between different age groups. Case studies have been pub-

ished using young children (Grelotti et al., 2005) suggesting thatith enough preparation and the careful design of simple experi-ental paradigms the imaging of younger cohorts may be possible.ock scanners can also help in acclimatizing people to the scanner

nvironment.Due to the demanding nature of fMRI it is not surprising that

here is a bias in the fMRI literature to investigate high-functioningndividuals. However, given reports that the majority of individu-ls with autism have a comorbid learning disability (Steffenburgt al., 2003) this is another imbalance in the literature that needsddressed. In addition to the practical difficulties of scanning anutistic group with additional learning disability, recruiting anppropriate comparison group is also a challenge. Due to the com-lex behavioural profile of autism which might present with bothnhanced ability in some aspects of cognition in addition to broadanging impairments, IQ may not be the most appropriate measureith which to ability match people with autism to a comparison

roup. Perhaps more stringent matching based on factors whichndicate background and environmental influences (such as edu-ation and socioeconomic status) combined with age (and possibleven pubertal status when investigating children and adolescents)nd gender would provide a better method than IQ measures whichan be an intrinsic element of the condition. Also, it is imperative
hat tasks continue to be designed in such a way that groups can beerformance matched. Resting state fMRI is an increasingly popu-
ar approach which may be more acceptable to individuals of lowognitive ability.

(x = 40, y = −42, z = −22) and left middle temporal gyrus (x = −52, y = −40, z = 0); (d)right inferior occipital gyrus (x = 30, y = −84, z = −14) and left fusiform gyrus (x = −20,y = −88, z = −14).

4.1.2. Image analysisThe methodology applied in the analysis of fMRI data varied

throughout the literature and it was not always possible to knowthe exact details of the contrast that was carried out or the exactimage analysis methods used which led to the reported results.The use of different statistical thresholds, including those that wereuncorrected for multiple comparisons, makes it difficult to directlycontrast findings between studies. Similarly, whilst the use of ROIapproaches was generally justified within each study, this addsfurther to the difficulty of synthesising data. Many studies alsoreported apparent differences on within group activation mapswhich were not statistically significant in between group contrasts.Other differences in imaging methodology also limit comparisonbetween studies: the use of random effects analyses and fixedeffects analyses may lead to different results, as does the use ofdifferent templates between studies, for example standard MNI
templates versus ones derived from the participants scans. Thechoice of template is a particularly important factor which mayinfluence results in populations in whom neuroanatomy is knownto differ from typical populations.


Table 8Auditory and language tasks. Abbreviations as in Table 2.




Mean IQ(sd)


Gaffrey et al.(2007)*

10:0 26.1 (10.5) 101.5(11.9)

8 autism, 2AS

DSM-IV, ADI,ADOS

Age, gender,handedness,Non-verbal IQ

Wordcategorisation(colours, tools andfeelings) via buttonpress; letterrecognition instrings ofnon-word letters asbaseline condition.

ASD > control: leftmedial frontal gyrus,middle temporal gyrus,lingual gyrus andcuneus, right lingualgyrus, middle occipitalgyrus, post centralgyrus, posteriorcingulate andprecuneus.

Gervais et al.(2004)* 5:0 25.8 (5.9) 81 (18.8) Autism DSM-IV, ADI Age, gender

1. Passive listeningto vocal sounds;silence as baseline.

1. Control > ASD: rightsuperior temporalsulcus region andbilateral superiortemporal gyrus.

2. Passive listeningto vocal sounds;environmentalsounds as baseline.

2. Control > ASD: rightmiddle temporal gyrusand bilateral superiortemporal sulcus region.

Gomot et al.(2006)*

12:0 13.5 (1.6) 116 (18) HFA DSM-IV, ADI Age, IQ, gender,handedness

Passive auditorystimulation whilstparticipantswatched a video.1. Novel sounds;standard sounds asbaseline.2. Deviant sounds;standard sounds asbaseline.

1. Control > ASD:bilateral inferiorparietal lobe andposterior superiortemporal gyrus, rightinferior and middlefrontal gyrus, leftanterior cingulate gyrusand right anteriorcerebellum.2. Control > ASD: leftanterior cingulate gyrus,left medial orbitofrontalregion and left inferiorfrontal gyrus.(uncorrected stats)

Gomot et al.(2008)*

12:0 13.5 (1.6) 116 (18) 9 HFA, 3 AS DSM-IV, ADI,AQ-adol

Age, IQ, gender, inscanner accuracy(but not RT)

Auditory oddballparadigm with 3conditions;standard, deviantand novel. Buttonpress to novelstimuli. Periods ofrest i.e. no sound,interspersed.

Novel vs standardsoundsASD > control: rightmiddle frontal gyrus,superior frontal gyrus,precentral gyrus,postcentral gyrus andinferior frontal gyrus,left inferior parietallobule and middlefrontal gyrus.Control > ASD: rightcaudatep < 0.001 uncorrected

Harris et al.(2006)

14:0 36 (12) 116 (8) 7 autism, 5AS, 2PDD-NOS

DSM-IV, ADI,ADOS

Age, gender,handedness, verbalIQ, taskperformance

Viewing concreteand abstract words.Perceptual task;upper/lower casediscrimination.Semantic task;positive/negativediscrimination.

Direct group comparisonnot performed. ASDgroup appeared to havegreater activation ofWernicke’s area and lessBroca’s area duringsemantic processing.Control group activationwas modulated by wordtype – an effect not seenin the ASD group.

Just et al.(2004)

17 Notreported

>80 HFA ADI, ADOS Age, IQ, gender,socioeconomicstatus

Sentencecomprehension(identifying theagent or recipientof the action);fixation baseline.

ASD > control: leftsuperior temporal gyrusControl > ASD: leftinferior frontal gyrusROI analysis restricted toleft superior temporalgyrus and left inferiorfrontal gyrusFunctional connectivityanalysisSee Table 8


Table 8 (Continued)





Kana et al.(2006)* 11:1 22.5 (8.8) 110.7 (9.2) Autism ADI, ADOS

Age, IQ, gender,socioeconomicstatus, taskperformance

1. Indicate viabutton presswhether alow-imagerysentence is true orfalse; fixationbaseline.

1. ASD > control: leftintraparietal sulcus,right superior parietallobe, bilateral cuneus,precuneus and lingualgyrus.

2. Indicate viabutton presswhether ahigh-imagerysentence is true orfalse; fixationbaseline.

2. Control > ASD: leftinferior frontal gyrus,left angular gyrus andleft middle frontalgyrus.Functional connectivityanalysisSee Table 8

Kleinhans et al.(2008a)*

14:0 24.1 (9.58) 98.14 (11.84) 8 autismdisorder, 3AS,3PDD-NOS

DSM-IV, ADI,ADOS

age 2 tasks of verbalfluency.1. Letter fluency:Participantsgenerate as manywords as possiblebeginning with theletter presented onscreen; self-pacedrepetition of theword ‘nothing’ isresponse to seeingthe word on screen(high levelbaseline); ‘rest’condition (lowlevel baseline).2. Categoryfluency:Participantsgenerate as manyitems as possiblefrom the categorypresented onscreen. High andlow level baselineconditions as task1.

1. Letter fluency vs highlevel baselineASD > control: rightinferior frontal lobe2. Category fluency vshigh level baselineControl > ASD: leftmiddle frontal lobe

Knaus et al.(2008)*

12:0 15.46 (2.8) 105.42(19.35)

ASD DSM-IV, ADI,ADOS (butnot all metcut off)

Age, gender,handedness

Language task withtwo conditions;Response namingcondition –participants wereshown a 3 worddescription andhad to indicate theitem described outof a choice of twopresented optionsvia button press;Letter judgementcondition – threeletters presentedand participantsindicated viabutton press if theywere in upper orlower case. Crosshair presented for4 seconds as thebeginning of eachblock.

Response naming vsletter judgementControl > ASD: near theposterior corpuscallosum/cingulate.ASD > control: rightinferior frontal gyrus,middle frontal gyrus,middle temporal gyrus,pre-central gyrus,orbito-frontal gyrus andsuperior parietal gyrus.Left inferior frontalgyrus, inferior temproalgyrus, fusiform gyrus,pre-central gyrus,posterior superiortemporal gyrus, medialsuperior/middle frontalgyrus, and medial parstriangularis/parsopercularis.


Table 8 (Continued)





Mason et al.(2008) 17:1 26.5 101.9 HFA ADI, ADOS

Age, IQ, gender,socioeconomicstatus, race

Read short scenariosand answered yes/nocomprehensionquestions via buttonpress.

ASD > control: right middletemporal gyrus, superiortemporal gyrus, angulargyrus and supramarginalgyrus.

Intentional, emotionaland physicalsentences; fixationbaseline.

Whilst the recruitment ofthese regions was modulatedby sentence type in the controlgroup (greater activationwhen making mentalisticinferences) the ASD grouprecruited these regions for allsentence types.Functional connectivityanalysisSee Table 8

Oktem et al.(2000)

9 12 (3.1) 76.78 AS DSM-IV Age, handedness Instructed to think overgeneral comprehensionquestions; rest asbaseline.

Control > ASD: frontal lobe.

Redcay andCourchesne(2008)*

12:0 2.9 (0.6) 11 autismdisorder, 1ASD

ADI-R, CARS Mental agematched group(MA) andchronological agematched group(CA)

Speech processing task.3 conditions; simpleforward speech,complex forwardspeech, backwardspeech and ‘rest’ i.e. noauditory stimulation

Response to forward speechvs ‘rest’MA control > ASD: regions ofbilateral frontal, temporal,parietal and occipital lobes,cerebellar cortex and rightcaudate.ASD > MA control: bilateralpostcentral gyriCA control > ASD: leftanterior cingulate, middlefrontal gyrus, superiortemporal gyrus, middletemporal gyrus and fusiformgyrus. Posterior regions ofbilateral parietal, extrastriateand cerebellar cortices.ASD > CA control: rightmiddle and inferior frontalgyri, insula and post centralgyrus.

Takeuchi et al.(2004)

8:2 11.2 (2.6) 110.4 (14.8) Autism DSM-III Age, IQ,handedness

Reading task involvingattribution of complexmental states

ASD > Control: rightprefrontal cortexp < 0.01 (uncorrected)

Wang et al.(2006)*

18:0 11.9 (2.8) 102 (18) Autism, AS ADI, ADOS Age, IQ, gender,handedness, socialresponsivenessscale

Participants listen toscenarios and indicatevia button press if thespeaker means whatthey say.1. Event knowledgeand prosodic cues;resting baseline.2. Event knowledgeonly; resting baseline.3. Prosodic cues only;resting baseline.

1. ASD > control: leftpre-central gyrus andbilateral inferior frontalgyrus.2. Control > ASD: left superiorfrontal gyrusASD > control: left pre andpost-central gyrus, superiortemporal gyrus, superiortemporal sulcus region andmedial temporal gyrus andright inferior frontal gyrus.3. ASD > control: left superiortemporal sulcus region andright temporal pole.

Wang et al.(2007)*

18:0 12.4 (2.9) 98 (17) ASD DSM-IV, ADI,ADOS

Age, IQ, gender,handedness, socialresponsivenessscale, taskperformance

Indicate via buttonpress if the speaker issincere or not.Storyboards andspeech stimuli.1. Participants asked topay attention; restingbaseline.2. Participants asked topay attention to facialexpression or tone ofvoice; resting baseline.3. All ironic scenarios;non-ironic scenarios.

1. Control > ASD: bilateralmedial prefrontal cortex,superior temporal gyrus andcerebellum.2. Control > ASD: superiortemporal gyrus, cerebellumand visual cortex.3. Control > ASD: bilateraltemporal regions and medialprefrontal cortex.



Table 9Group comparisons of regions with significantly elevated likelihood of activation in auditory and language tasks. Brain areas activated from the ALE analysis (pFDR < 0.05 anda minimum cluster size of 200 voxels). C/A: children/adolescents.


x y z

ASD > C: All Precentral Gyrus R 6 1704 36 −10 54 1.13Declive L 720 −18 −78 −18 1.50Frontal Sub-Gyral L 6 392 −20 2 52 1.20Posterior Cingulate L 30 264 0 −50 14 1.07

ASD > C; C/A Precentral Gyrus R 6 816 34 −8 54 0.99

ASD > C; Adults Declive L 896 −18 −78 −18 1.50Posterior Cingulate L 30 384 0 −50 14 1.07Declive R 336 20 −70 −14 0.90Precuneus R 19 336 12 −78 38 0.82

C > ASD; All Superior Temporal Gyrus R 22 1344 56 −8 −2 1.90Superior Temporal Gyrus L 41 1144 −52 −20 10 1.60Pyramis R 224 10 −78 −24 1.21Cingulate Gyrus L 32 224 −14 18 24 1.31

C > ASD; C/A Superior Temporal Gyrus L 41 1048 −52 −20 10 1.59Superior Temporal Gyrus R 22 608 56 −8 −2 1.66Pyramis R 224 10 −78 −24 1.20

4

itrfgimtdo

although not part of the diagnostic criteria, are regarded as a

Faga(

Cingulate Gyrus L 32

C > ASD; Adults Cingulate Gyrus L 32

.1.3. In-scanner behavioural performanceIndividuals with ASD and controls were generally well matched

n terms of behavioural performance in the scanner, giving strengtho the idea that differences seen in activation are not simplyelated to performance differences. It is somewhat unusual that soew studies report performance differences between the groups,iven that many of the domains under examination are onesn which performance differences in a non-scanning environ-

ent have been noted. Tasks used in a scanner may be simpler
han those used elsewhere, as they need to be able to be con-ucted in such an unusual environment. It is notable that tasksf complex social cognition were not behaviourally matched,
ig. 6. Complex social tasks. Whole group analysis. ALE maps (pFDR < 0.05) are superimposreas of greater probability of activation in ASD subjects compared to controls in a clusteyrus (x = −20, y = −22, z = 64) and left postcentral gyrus (x = −40, y = −22, z = 54); (c) leftctivation likelihood estimate maps in clusters centred at (d) right superior temporal gyrf) left inferior parietal lobe (x = −54, y = −56, z = 40).

224 −14 18 24 1.30

536 −4 22 34 1.23

possibly due to the inherent need for complexity within thesetasks.

4.2. Task related activation differences

4.2.1. Motor tasksMotor coordination impairments were included in Asperger’s

early descriptions of autistic psychopathy (Asperger, 1944) and,

major feature of ASD (Fournier et al., 2010). Both hypo- and hyper-activation were seen in the meta-analysis, including in ‘traditional’motor regions such as the anterior cerebellum, the precentral gyrus

ed on slices from grey matter template in Talairach space. The top panel illustratesr centred at (a) right inferior frontal gyrus (x = 40, y = 22, z = 10); (b) left precentral

superior temporal gyrus (x = −48, y = −32, z = 4). The bottom panel shows C > ASDus (x = 56, y = −8, z = −2); (e) left superior temporal gyrus (x = −52, y = −20, z = 10);


Table 10Basic social tasks. Abbreviations as in Table 2.





Ashwin et al.(2007)*

13:0 31.2 (9.1) 108.6 (17.1) 1 HFA, 12 AS Clinicaldiagnosis, AQ


Button press inresponse to stimuli.Face condition(including highfear, low fear andneutral faces);scrambled facestimuli as baseline.

Control > ASD: leftamygdala andorbitofrontal cortex.ASD > control: rightanterior cingulatecortex and bilateralsuperior temporalcortex.In controls, intensity offear modulated activityin bilateral amygdala,fusifrom gyrus, rightmedial prefrontal cortexand superior temporalsulcus region, an effectnot seen in the ASDgroup.

Bird et al.(2006)* 14:2 33.3 (11.5) 119 (14) 1 autism,

15 ASDSM-IV,ADOS, AQ

Age, IQ, gender,taskperformance

1. Localiser task:passive viewing ofneutral faces;passive viewing ofhouses; fixationbaseline. (fixationcross over ‘eyeregion’)

1. No between groupdifferences in anycontrast.

2. Same/differentdiscrimination inhorizontal orvertical plane viabutton press; pairsof faces and housespresented.

2. Attend to houses vsnot attendNo between groupdifferences.Attend to faces vs notattendControl > ASD group:left fusiform gyrus.Connectivity analysisSee Table 8

Bookheimeret al. (2008)

12:0 11.3 (4) Not reported 8 autism, 2AS,2PDD-NOS

ADI-R, ADOS,SCQ

Age, gender Matching task withthree conditions;neutral faces, allupright; neutralfaces, targetinverted; shapematching task as ahigh level baselinecondition.

Upright condition vsbaselineControl > ASD: Leftprefrontal cortexInverted condition vsbaselineControl > ASD: leftprefrontal cortexASD > control: bilateralprecuneusFunctionally selectedROI analysis. No directgroup comparison atwhole brain levelreported. Precuneus ROIapplied post hoc.

Corbett et al.(2009)*

12:0 9.01 (1.6) 90.71 (13.82) 12 HFA DSM-IV, ADIand ADOSwherepossible

Age, performancein scanner

Participantsindicate via buttonpress;1. Face emotionmatch2. Face identitymatch3. Object categorymatch – notreported on4. Pattern match –baseline

Emotion > patternmatchControl > ASD: RightfusiformASD > control: Leftsuperior parietal lobe,precentral gyrus,middle frontal gyrusNo significant betweengroup differences inidentity matchconditionFusiform and amygdalavolume of interestanalysisIdentity > pattern matchControl > ASD: leftamygdala and rightfusiform


Table 10 (Continued)





Critchley et al.(2000)*

9:0 37 (7) 102 (15) Autism, AS ICD-10, ADI Age, IQ Explicit emotiontask: indicate viabutton press if facestimuli arehappy/angry orneutralImplicit emotiontask: indicategender via buttonpress (samestimuli).

Emotion vs neutralstimuliASD > control: leftsuperior temporal gyrusand peristriate visualcortex.Control > ASD: rightfusiform cortex.Significant group xcondition interaction incerebellar vermis, leftlateral cerebellum,striatum, insula andamygdalohippocampaljunction, and middletemporal gyrus.

Dalton et al.(2005)*

1. 11:0 1. 15.9 (4.7) 1. 94 (19.5) 1. Autism, AS 1. DSM-IV,ADI

1. Age, gender 1. Indicate viabutton press ifstimuli emotionalor neutral. Happy,fear, anger andneutral face stimulipresented (halfquarter turned, halffacing ahead);resting baseline.

1. Control > ASD: bilateralfusiform, occipital gyrusand middle frontal gyrus.ASD > control: leftamygdala andorbitofrontal gyrus.Differences did not relateto orientation oremotional content.Fixation on eyes correlatedwith level of activity in leftamygdala and rightanterior fusiform gyrus inASD group, effect not seenin controls.

2. 16:0 2. 14.5 (4.6) 2. 92.1 (27.7) 2. Autism, AS 2. DSM-IV,ADI

2. Age, gender 2. Discriminatefamiliar andunfamiliar faces viabutton press;resting baseline.

2. Control > ASD: bilateralfusiform, left anteriormedial cortex, leftposterior lateral cortex,right occipital cortex.Greater activation in rightoccipital and fusiformgryus in control group inresponse to familiarity,effect not seen in ASDgroup. Fixation on eyescorrelated with level ofactivity in right amygdalaand right anterior fusiformgyrus in ASD group, effectnot seen in controls.

Deeley et al.(2007)*

9:0 34 (10) 114 (12) AS DSM-IV,ICD-10, ADI,ADOS

IQ, gender,handedness, taskperformance

Genderdiscrimination viabutton press.Neutral faces;emotional faces(high and lowintensity); fixationbaseline.1. Fear vs baseline2. Disgust vsbaseline3. Happy vsbaseline4. Sad vs baseline5. Neutral faces(from eachemotion condition)vs baseline

1. Control > ASD: rightfusiform gyrus andcerebellum and left preand postcentral gyrus.2. Control > ASD: leftfusiform gyrus, lingualgyrus and cerebellum.3. Control > ASD: leftfusiform gyrus, lingualgyrus and cerebellum.4. Control > ASD: bilateralfusiform gyrus, lingualgyrus and cerebellumand left inferior occipitalgyrus.5. Control > ASD:fusiform, lingual,occipital cortices andcerebellum.Intensity of emotion hadan effect on brainactivation at a trend levelin both groups whichvaried with each emotion.







Dichter andBelger (2007) 16:1 22.9 (5.2) 105 (18.6) 14 HFA, 3

ASDSM-IV,ADI, ADOS

Age, IQ,gender,handedness,education,taskperformance

1. Indicate viabutton pressdirection of centralarrow that wascongruent(condition1) orincongruent(condition2) withdirection of flankerarrows.

1. No significant groupdifference.

2. Indicated viabutton pressdirection of eyegaze in central facethat was congruent(condition1) orincongruent(condition2) withdirection of eyegaze in flankerfaces.

2. Control > ASD: bilateraldorsolateral prefrontalcortex, right inferiorfrontal/anterior insulacortex, anterior cingulateand bilateral intraparietalsulcus. (ROI analysis)

Dichter andBelger(2008)

12:0 23.2 (5.8) 106.9 (19.2) HFA/AS DSM-IV,ADOS, ADI

Age, IQ, in scanneraccuracy (but notRT)

Participantscompleted a 2AFCtask, indicatingarrow direction viabutton press.3 conditions werepresented; neutralarrows, congruentand incongruenttrails.Each trail waspreceded with ahigh or low arousalimage presentedfor 200 ms.

No significant betweengroup differences werereported.Within group maps weresimilar in the ASD andcontrol groups when theincongruent trails werepreceded by low arousalimages. When preceded byhigh arousal images theASD group did notmodulate activity in rightlateral midfrontal cortexas seen in the controlgroup.

Freitag et al.(2008)*

13:2 17.5 (3.5) 101.2 (21.2) ASD DSM-IV, ADI,ADOS

Age, IQ, gender,handedness

Indicate via buttonpress if stimulus isbiological motionor scrambled.Point-lightwalkers; scrambledversion; fixationbaseline.

1. All motion stimuli vsfixation: no groupdifferences.2. Biological motion vsscrambled Control > ASD:right middle temporalgyrus, medial and middlefrontal gyrus and leftanterior superiortemporal gyrus, fusiformgyrus, and bilateral postcentral gyrus and inferiorparietal lobule.(uncorrected stats.)

Greimel et al.(2009)*

15:0 14.9 (1.6) 109.9 (17.3) 12 AS, 3 HFA DSM-IV,ICD-10,ADOS, ADI-R,SCQ

Age, IQ, gender,handedness

Participants arepresented withface stimuli andindicate theirresponse viabutton press in 3conditions. Othercondition; judgeface stimuli ashappy, sad orneutral. Selfcondition; judgeown response toface as happy, sador neutral. Highlevel baseline;judge neutral facesas thin, average orwide.

Other condition vs highlevel baselineControl > ASD: leftsuperior occipital gyrus,left cuneus, right anteriorfusiform gyrus.Self condition vs highlevel baselineControl > ASD: leftinferior frontal gyrus.IFG and FG results werepart of an ROI analysis.







Grèzes et al.(2009)*

10:2 26.6 (10.4) 102 (20.6) 10 AS, 2 HFA DSM-IV Age, IQ,behaviouralperformance inscanner

Oddball paradigm;button press toupside downstimuli. Blankscreen baseline.1. Fearful dynamicbody stimuli2. Neutral dynamicbody stimuli3. Fearful staticbody stimuli4. Neutral staticbody stimuli

Dynamic > static(1 + 2 > 3 + 4)Control > ASD: Righttemporo-parietaljunction, STG, middleSTS, inferior temporalgyrus, medial superiorfrontal gyrus, precentralgyrus, precuneus,fusiformgyrus/cerebellum. Leftinferior temporal gyrusand inferior frontal gyrus.Fear > neutral(1 + 3 > 2 + 4)Control > ASD: Rightinferior frontal gyrus,precentral gyrus, inferiortemporal gyrus,amygdala.ASD > controls: Leftmedial anterior superiorfrontal gyrusInteraction betweenfearful body expressionand dynamic information(1–3 > 2–4)Controls > ASD: Rightprecuneus, STS, IFG,middle cingulate cortex,lingual gyrusASD > controls: Righttemporal gyrus. Lefttemporal pole/insula andtemporo-parietaljunctionA threshold of p < 0.001(uncorrected) was applied.

Hadjikhaniet al. (2004a)

11:0 36 (12) 119 (8) Autism, AS,PDD-NOS

DSM-IV,ADI-R, ADOS,

IQ, gender Passive viewing ofneutral faces withfixation cross overeye region;scrambled faces asbaseline.

No significant differencesbetween ASD and controlgroup when activationwithin ROI’s compared.

Hadjikhaniet al. (2007)

8:2 34 (11) 124 (10) 6 autism, 3AS, 1PDD-NOS

ADI-R, ADOS Age Passive viewing ofneutral faces withfixation cross overeye region;scrambled faces asbaseline.

Control > ASD: rightsuperior temporal sulcusregion, somatosensoryand premotor cortex,inferior frontal cortexand amygdala.(ROI analysis)Fusiform Face Areaactivation in ASD groupnot significantly differentfrom controls.

Hadjikhaniet al. (2009)*

9:3 30 (11) 126 (10) 8 autism, 3AS,1PDD-NOSthreeexcluded dueto movement

ADI, ADOS Passive viewing ofstatic body images(faces blurred)presented in blocksexpressing fearfuland neutralemotion

Processing fearful bodyimagesControl > ASD: bilateralcolliculus, pulvinar,amygdala, extrastriatecortex, ventraltemporal-occipitalcortex, nucleusaccumbens, anteriorinsula, putamen, motorand premotor cortex andinferior frontal cortex.No between groupdifferences apparent inthe neutral condition.







Herringtonet al. (2007)*

10:0 27.6 (7.1) 109 AS DSM-IV,ICD-10


Button press toindicate directionof movement.Point-lightwalkers; scrambledversion; fixationbaseline.

1. Scrambled movement vsfixationControl > ASD: right superiortemporal gyrus and angulargyrus.2. Walkers vs fixationControl > ASD: bilateralcerebellum, fusiform gyrus,middle temporal gyrus,middle occipital gyrus,cuneus, right inferiortemporal gyrus and inferioroccipital gyrus, and leftsuperior temporal gyrus,inferior parietal lobe,angular gyrus, precuneusand precentral gyrus.

Hubl et al.(2003)

7:0 27.7 (7.8) 98 (17) Autism ICD-10, ADI,ADOS

Age, IQ, gender,task performanceaccuracy (not RT)

1. Indetify targetvia button press.Happy, sad, angryand neutral facespresented. Targethappy (explicitemotioncondition); targetfemale (implicitemotioncondition);scrambled faces asbaseline condition.2. Colour counting;shape countingwithin a mosaic;and a restcondition.

Values from ROIs wereextracted from each conditionand investigated in an ANOVAfor interactions withdiagnosis, task, region andhemisphere.Activations for each task weredifferent between ASD andcontrols.Different tasks activateddifferent regions.The ASD and control groupsdiffer with respect to thedifference in activationscaused by each task indifferent regions.

Humphreyset al. (2008)* 13:0 27 (10)

VIQ 103PIQ 106 Autism ADI-R, ADOS Age, gender

1. Blocks of linedrawings of faces,buildings, objectsand patternspresented. Pressbutton whenrepetition occurs

1. Control > Autism: bilateralfusiform face area for facesand right fusiform face areafor objects. Selectivity forfaces in the left fusiform facearea.

2. Passive viewingof blocks of moviesof faces, buildings,open fields andobjects presented.

2. Control > Autism: bilateralfusiform face area for faces.Selectivty for faces inbilateral fusiform face areaAutism > ControlBilateral fusiform face areafor scenes.In experiment 2 greaterinter-individual variabilitywas seen in the fusiform facearea in the autism groupthan the controls. This effectwas not seen in other ROIs –lateral occipital cortex andcollateral sulcusROI analysis

Kennedy et al.(2006)*

12:0 26.5 (12.8) 101.6 (15.2) 6 autism, 6AS

ADI-R, ADOS Age Counting Strooptask withemotional, neutraland number words.Participants askedto count number ofwords on thescreen. Baselinefixation cross

Number vs restASD > Control: rightsupramarginal gyrus, rightprecuneus, bilateral inferiorparietal lobule, rightsuperior frontal gyrus, leftmedial frontal/anteriorcingulateEmotional vs neutralControl > ASD: right medialorbitofrontal cortex, rightmiddle occipital gyrusFunctional connectivityanalysisSee Table 8







Kleinhans et al.(2008b)*

19:0 23.5 (7.8) 106.7 (15.7) 8 autism, 9AS, 2PDD-NOS

DSM-IV,ADI-R, ADOS


1-back withneutral faces;1-back with housesas baseline.

No significant groupdifferences.Functional connectivityanalysisSee Table 8

Kleinhans et al.(2009)*

19 21.9 (5.9) 107 (13.8) ASD DSM-IV,ADOS, ADI

Age, IQ, in scanneraccuracy (but notRT)

Participantscomplete a 1 backworking memorytask with blocks ofstimuli; uprightneutral faces,inverted faces,houses and fixationbaseline condition.Two runs of thetask werecompleted, eachwith differentstimuli.

Upright neutral faces vsfixation Run 1No significant groupdifferencesUpright neutral faces vsfixation Run 2ASD > control, leftamygdala and rightfusiform gyrusFaces vs fixationRun1 > Run2Control > ASD: bilateralamygdalaSmall volume correctionsapplied.




Control > ASD: leftinferior prefrontal andright posterior temporalcortex.Different location of theFusiform Face Area withinthe fusiform gyrus in ASDgroup compared tocontrols.Functional connectivityanalysisSee Table 8

Loveland et al.(2008)

4:1 18.25 (15.9) 112.6 (15.3) Autism DSM-IV,ADI-R, ADOS

Age, handedness Emotional facesand voicespresented incongruent andincongruent pairs.Participants askedto judge emotionalcongruence(condition ofinterest) or gendercongruence(baseline).

Control > ASD: bilaterallingual gyrus, bilateralcuneus, right middlefrontal gyrus, leftparahippocampal gyrus,left fusiform.

Ogai et al.(2003) 5 21.8 (5.9) 112.4 (10.5) HFA DSM-IV

Age, IQ,socioeconomicstatus, education,emotion label taskperformance

Instructed to thinkabout the emotionbeing expressed.1. Happy faces;neutral faces

1. No significantdifference betweengroups.

2. Fear faces;neutral faces

2. Control > ASD: leftmiddle frontal gyrus.

3. Disgust faces;neutral faces

3. Control > ASD: leftinsula, left inferiorfrontal gyrus and leftputamen.

Pelphrey et al.(2007)* 6:2 24.5 (11.5) 120 (9) Autism ADI, ADOS


Button press inresponse to facestimuli. Staticneutral; staticemotional (angerand fear); dynamicneutral (identitymorph); dynamicemotional(emotion morph)events. Fixationbaseline.

1. Dynamic emotionalcondition vs fixationbaselineControl > ASD: rightamygdala and superiorfrontal gyrus, leftfusiform gyrus andmedial frontal gyrus andbilateral middletemporal gyrus.







2. Dynamic emotionalcondition vs staticemotionGroup x conditioncontrast: control groupmodulate activity inamygdala, superiortemporal sulcus regionand fusiform gyrus, aneffect not seen in ASDgroup.3. Dynamic neutralcondition vs staticneutralNo group differences.4. Static emotioncondition vs fixationbaselineASD > control: superiortemporal sulcus region.(No amygdala or fusiformgyrus differences) (ROIanalysis)

Pelphrey et al.(2005)

9:1 23.2 (9.9) 107 (16) Autism ADI, ADOS Age, IQ, taskperformance

Button press inresponse to shiftsin eye gaze thatwere congruentwith the location ofa visual stimulus;incongruent trials.

Group x conditioncontrast: posteriorsuperior temporal sulcusregion activity modulatedby congruence in controlgroup but not in ASDgroup. Inferior frontalgyrus and insular cortexmodulated by congruencein ASD group but notcontrol group.

Pierce et al.(2001)

7:0 29.5 (8) 83.7 (10.9) Autism DSM-IV,ADI-R, ADOS,CARS


Button press inresponse to target.Neutral faceperception; shapeperception.

Control > ASD: bilateralfusiform and leftamygdala.(ROI analysis, butfindings supported bywhole brain withingroup maps.)

Pierce et al.(2004)

8:0 27.1 (9.2) 80.3 (17.7) Autism DSM-IV, ADI,ADOS


Genderdiscrimination viabutton press withfamiliar andstranger faces;fixation baseline.

No significant betweengroup differences.

Pierce andRedcay(2008)

9:2 9.9 (2.1) 108.5 (12.6) 9 autism, 1AS, 1PDD-NOS

ADI, ADOS Age, gender,handedness, taskperformance

1-back task. Buttonpress if stimulirepeated. Familiaradult; strangeradult; familiarchild; strangerchild; objects;fixation baseline.

No between groupdifferences reported inwhole brain analysis.Stranger faces vs baselineControl > ASD: leftfusifrom.Familiar children vsbaselineControl > ASD: posteriorcingulate.(ROI analysis)

Piggot et al.(2004) 14:0 13.1 (2.5) 112 (15.9) 7 autism, 7

ASDSM-IV, ADI,ADOS

Age, IQ,handedness,socioeconomicstatus, taskperformanceaccuracy (but notRT)

1. Match target byemotion to one oftwo responseoptions. Fearful,surprised andangry face stimuli;shape matching asbaseline.

1. Control > ASD: averagefusiform gyrus.(ROI analysis.)

2. Label emotiongiven two textchoices. Fearful,surprised andangry face stimuli;shape matching asbaseline.

2. No significant groupdifferences.







Schultz et al.(2000)

14:0 23.8 (12.4) 109.1 (19.5) 8 HFA, 6 AS ICD-10,ADI-R, ADOS,Vineland


Same/differentdiscrimination viabutton press.Neutral faces,objects; patterns.

Faces vs patternsControl > ASD: rightfusiform gyrusASD > control: bilateralinferior temporal gyrusactivation.No differences betweengroups in object task.(ROI analysis)

Wang et al.(2004)* 12:0 12.2 (4.8)

Notreported

Autism, AS,PDD-NOS

ADI, ADOSAge, gender,handedness,nonverballanguage age,taskperformance

1. Match target byemotion to one oftwo responseoptions. Fearfuland angry facestimuli; shapematching asbaseline.

1. Control > ASD: bilateralfusiform gyrus.ASD > control: precuneus.

2. Label emotiongiven two textchoices. Fearful andangry face stimuli;shape matching asbaseline.

2. No significant groupdifferences.(ROI analysis)

Wicker et al.(2008)*

11:1 27 (11) 81.3 8 autism, 4AS

DSM-IV Age Explicit angry –happy faceemotionjudgements;baseline explicitgender judgement.Actors gaze either

Control > ASDright temporal-parietaljunction, right inferiorfrontal gyrus and medialsuperior frontal gyrusEffective connecitivityanalysisSee Table 8

arso2sc

rwmi

TGc


nd the basal ganglia, as well as other regions, such as the supe-ior and inferior parietal lobules which may relate to attentionalystems. Cerebellar abnormalities tended to be found in studiesf simple motor tasks (Allen and Courchesne, 2003; Allen et al.,004), whereas more complex sequence learning tasks tend tohow differences in cortical-subcortical networks as opposed to theerebellum (Muller et al., 2003, 2001, 2004).

It is important to consider these findings when interpreting
esults from other functional imaging studies in this population,here in most cases, a motor response via button press is theethod used to gather behavioural performance data. Tasks requir-
ng a motor response could disrupt cortico-cerebellar networks

able 11roup comparisons of regions with significantly elevated likelihood of activation in basic

luster size of 200 voxels). C/A: children/adolescents.

Comparison – age group Brain region BA V

ASD > C; All Superior Temporal Gyrus L 22

Superior Temporal Gyrus R 22

ASD > C; C/A No clusters found.

ASD > C; Adults Superior Temporal Gyrus L 41

Superior Temporal Gyrus R 22

C > ASD: All Fusiform Gyrus L 18 2Inferior Occipital Gyrus R 18 1Culmen R

Middle Temporal Gyrus L 22

C > ASD; C/A Cuneus R 17

Parahippocampal Gyrus R 36

C > ASD; Adults Fusiform Gyrus L 18 2Inferior Occipital Gyrus R 18

direct or averted

and alter connectivity patterns as a consequence of the responserequired rather than the task.

4.2.2. Visual processing tasksThe ALE meta-analysis revealed greater activation in control

subjects than in individuals with ASD in visual areas, such as thelingual and occipital gyri, which, given the lack of a behaviouraldifference between the groups, may reflect more efficient pro-
cessing of visual stimuli in these regions in individuals with ASD.This is consistent with the idea that individuals with ASD showenhanced perceptual systems (Samson et al., 2011; Soulieres et al.,2011).
social tasks. Brain areas activated from the ALE analysis (pFDR < 0.05 and a minimum

olume (mm3) Talairach ALE (10−2)

x y z

984 −60 −28 4 1.14360 54 8 4 0.89

392 −54 −24 6 0.92320 54 8 4 0.89

424 −20 −88 −14 2.73080 30 −84 −14 3.14248 40 −42 −22 1.99248 −52 −40 0 1.85

336 14 −82 8 0.91304 26 −36 −14 0.85

112 −20 −88 −14 2.73944 30 −84 −14 3.14


Table 12‘Complex’ social cognition tasks. Abbreviations as in Table 2.





Baron-Cohenet al. (1999)*

4:2 26.3 (2.1) 108.5 (10.5) Autism, AS DSM-IV,ICD-10

Age, IQ,handedness,socioeco-nomic status,education

Two alternate forced choicedecision via button press on themental state portrayed in eyestimuli; gender discriminationof same stimuli as baseline.

Control > ASD: leftinsula, inferior frontalgyus and right insula.ASD > control: bilateralsuperior temporalgyrus.

Chiu et al.(2008)

12:0 16.5 (3.3) 103 (18) Autism, AS,PDD-NOS

DSM-IV,ADI, ADOS

Age, IQ,gender, taskperformance

Multiround trust game –investor (control subject) isgiven an amount of money andchooses to send a proportion ofthis to trustee (ASD subject).This is tripled and trustee thenrepays a proportion of thetripled amount.1. ‘other’ condition; at the timeof controls investment.2. ‘self’ condition; when ASDsubject returns proportion ofthe money.

1. no significantdifference2. Control > ASD:middle cingulate.Reduced cingulateresponse in ASD groupcorrelates with total,social andcommunication ADIscores. Controls activatecingulate in both ‘self’and ‘other’ conditions,unless they were playingwith a computer inwhich case they activatein a similar fashion tothat seen in ASDsubjects here –suggesting dysfunctionin self-referentialprocessing in ASD.(ROI analysis)

Dapretto et al.(2006)*

9:1 12.05 (2.5) 96.4 (18.3) HFA ADOS, ADI Age, IQ Imitation of emotionalexpressions; observation ofemotional expressions; fixationbaseline.

Imitation vs restControl > ASD: bilateralinferior frontal gyrus,insula, periamygdaloidregions, ventralstriatum and thalamus.ASD > control: leftanterior parietal andright visual associationareas.Observation vs restControl > ASD: bilateralinferior frontal gyrus.

Gilbert et al.(2009)

14:2 32 (7.7) VIQ 117 (13.7)PIQ 115 (14.3)

ASD Clinical dxand ADOS(but not allmetcriteria)

Age,handedness,IQ

Spatial task – participantsfollow outline of shape viabutton press.Alphabet task – participantsindicate if letters were made upof only straight lines orincluded curvesBoth tasks have two conditions:stimulus orientated responses –where shapes or letters arevisible; stimulus independentresponses – where the task iscontinued without visual cuesAlso the pace of stimulipresentation varied and theparticipants were asked a) if theperson controlling the pace ofthe presentation was beinghelpful or not (a mentalizingcondition) or b) if thepresentation was fast or slow(non-mentalizing condition)Same-function comparisons –mentalizing vs mentalizing andnon-mentalizing vsnon-mentalizingCross function comparisons –mentalizing vs non-mentalizing

Multi-voxel similarityanalysis within medialrostral prefrontalcortex:Control group showdifferent distribution ofactivation when samefunction comparisonswere compared tocross functioncomparisons whereasASD group do not.Controls showedsimilar distributions ofactivation duringmentalizing conditionregardless of taskformat (spatial oralphabet) and duringnon-mentalizingcondition regardless oftask format. ASD groupdid not show thispattern indicatinggreater functionalspecialization incontrols.







Greimel et al.(2009)*

15:0 14.9 (1.6) 109.9 (17.3) 12 AS, 3HFA

DSM-IV,ICD-10,ADOS,ADI-R, SCQ

Age, IQ,gender,handedness

Participants are presented withface stimuli and indicate theirresponse via button press in 3conditions. Other condition;judge face stimuli as happy, sador neutral. Self condition; judgeown response to face as happy,sad or neutral. High levelbaseline; judge neutral faces asthin, average or wide.

Other condition vs highlevel baselineControl > ASD: leftsuperior occipitalgyrus, left cuneus, rightanterior fusiformgyrus. Self condition vshigh level baselineControl > ASD: leftinferior frontal gyrus.IFG and FG results werepart of an ROI analysis.

Kana et al.(2009)*

10:2 24.6 (6.9) 104.3 (14.4) Autism ADOS, ADI Age, IQ,socioeco-nomic status,race andbehaviouralperformancein scanner

4AFC of text to describe ananimation of geometric shapes.Event related design with 3conditions and fixation baseline.ToM interactionGoal Directed movementRandom movement

ToM > RandomControl > ASD: leftsuperior medial frontalcortex, anteriorparacingulate cortexand inferior orbitalfrontal gyrus. Bilateralanterior cingulategyrus.ASD > control: rightanterior superiortemporal gyrusp < 0.005 uncorrected

Kennedy andCourchesne(2008)

13:0 26.9 (12.3) 101.7 (14.6) 6 autism, 6AS,1PDD-NOS

ADI-R,ADOS

Age, IQ,gender,handedness

Participants were provided withstatements on which they maketrue/false judgments.Mental condition: Statementswere about themselves (SELFcondition) or a close otherperson (OTHER condition) andreferred to psychologicalpersonality traits (INTERNALcondition) or observableexternal characteristics orbehaviours (EXTERNALcondition).High level baseline condition:participants presented withmaths equations and indicatedwhether the answer was true orfalse for that sum.Low level baseline; visualfixation.

All mental judgementconditions vs high levelbaseline:Control > ASD: ventralmedial prefrontalcortex/ventral anteriorcingulate cortex.All mental judgementsvs low level baseline:Control > ASD:dorsomedial prefrontalcortex,retrosplenial/posteriorcingulate cortex andleft angular gyrusHigh level baseline vslow level baselineControl > ASD: ventralmedial prefrontalcortex/ventral anteriorcingulate cortex.ROI analysis of defaultnetwork.

Mason et al.(2008)

17:1 26.5 101.9 HFA ADI, ADOS Age, IQ,gender,socioeco-nomic status,race

Read short scenarios andanswered yes/nocomprehension questions viabutton press.Intentional, emotional andphysical sentences; fixationbaseline.

ASD > control: rightmiddle temporal gyrus,superior temporalgyrus, angular gyrusand supramarginalgyrus.Whilst the recruitmentof these regions wasmodulated by sentencetype in the control group(greater activationwhen makingmentalistic inferences)the ASD group recruitedthese regions for allsentence types.Functional connectivityanalysisSee Table 8







Pinkham et al.(2008)*

12:0 24.08 (5.71) 110 (10.91) HFA DSM-IV,ADI, ADOS

Age, gender,handedness,verbal IQ

Indicate judgement of facestimuli via button press.Trustworthiness; age; fixationbaseline.

1. Trust vs baselineControl > ASD: rightamygdala and FusiformFace Area and leftventerolateralprefrontal cortex.2. Age vs baselineASD > control: leftsuperior temporalsulcus region, rightventerolateralprefrontal cortex.3. Trust vs ageComparison of withingroup maps: amygdala,superior temporal sulcusregion andventerolateral prefrontalcortex activitymodulated in controlgroup, effect not seen inASD group.(ROI analysis.)

Silani et al.(2008)*

13:2 36.6 (11.7) 117.6 (13.5) HFA/AS DSM-IV,ADOS

Age, IQ,gender andbehaviouralperformancein scanner

Internal task; participants ratedpictures (unpleasant, neutraland pleasant) on the emotioninvoked by the picture (positiveto negative)External task; participants ratedpictures (unpleasant, neutraland pleasant) on colour balancein the picture (black to white)

Internal task > externaltaskControl > ASD: medialprefrontal cortex, ACC,precuneus, lefttemporal lobe andcerebellumASD > control: parietaland occipital cortexUnpleasant > pleasantstimuliControls > ASD: Rightcerebellum. Leftinferior orbitofrontalcortexASD > controls: Rightmiddle occipital cortexand calcarineInternal (unpleasant –neutral) > external(unpleasant – neutral)Control > ASD: Rightcorpus callosump < 0.001 uncorrected

Takeuchi et al.(2004)

8:2 11.2 (2.6) 110.4 (14.8) Autism DSM-III Age, IQ,handedness

Reading task involvingattribution of complex mentalstates

ASD > Control: rightprefrontal cortexp < 0.01 (uncorrected)

Uddin et al.(2008)

12:0 13.19 (2.61) 116 (14) HFA ADI, ADOS(except onepartici-pant)

Age, IQ,gender,handedness,in scannerperformance

Faces presented morphing from100% self to 100% other plus ascrambled image, plus fixationbaseline.Participants indicated viabutton press whether the facewas self or other.

No significant betweengroup differences inWBA.Whilst both groupsactivated right inferiorfrontal gyrus whenviewing ‘self’ images,only the control groupactivated this regionwhen viewing ‘other’images. This findingreached statisticalsignificance when afunctionally derived ROIwas applied to the rightfrontal cortex.







Wang et al.(2006)*

18:0 11.9 (2.8) 102 (18) Autism, AS ADI, ADOS Age, IQ,gender,handedness,socialresponsive-nessscale

Participants listen to scenariosand indicate via button press ifthe speaker means what theysay.1. Event knowledge andprosodic cues; resting baseline.2. Event knowledge only;resting baseline.3. Prosodic cues only; restingbaseline.

1. ASD > control: leftpre-central gyrus andbilateral inferior frontalgyrus.2. Control > ASD: leftsuperior frontal gyrusASD > control: left preand post-central gyrus,superior temporal gyrus,superior temporal sulcusregion and medialtemporal gyrus and rightinferior frontal gyrus.3. ASD > control: leftsuperior temporal sulcusregion and righttemporal pole.

Wang et al.(2007)*

18:0 12.4 (2.9) 98 (17) ASD DSM-IV,ADI, ADOS

Age, IQ,gender,handedness,socialresponsive-ness scale,taskperformance

Indicate via button press if thespeaker is sincere or not.Storyboards and speech stimuli.1. Participants asked to payattention; resting baseline.2. Participants asked to payattention to facial expression ortone of voice; resting baseline.3. All ironic scenarios;non-ironic scenarios.

1. Control > ASD: bilateralmedial prefrontal cortex,superior temporal gyrusand cerebellum.2. Control > ASD: superiortemporal gyrus,cerebellum and visualcortex.3. Control > ASD: bilateraltemporal regions andmedial prefrontal cortex.

Williams et al.(2006)*

16:0 15.4 (2.24) 100.4 (21.7) ASD ADI, ADOS Age, IQ,gender

Passive viewing task. Hand withindex or middle finger raised;hand stimuli with cross onmiddle or index finger; cross onleft or right side of screen;resting baseline.Hand with index or middlefinger raised (imitation); handstimuli with cross on middle orindex finger (execution); crosson left or right side of screen(execution); resting baseline.

Imitation vs restControl > ASD; rightfusiform gyrus, middleoccipital gyrus, lingualgyrus, middle temporalgyrus and bilateralinferior parietal lobe, andgreater activation in rightparahippocampal gyrusand cingulate gyrus, leftuncus, precentral gyrus,claustrum, middle frontalgyrus, middle occipitalgyrus and bilateralsuperior temporal gyrus.Imitation vs executionControl > ASD group hadless activation in rightprecuneus, left anteriorcingulate and inferiorparietal lobe.ASD > control; leftsuperior parietal lobeand bilateral precentralgyrus and middle frontalgyrus.ROI of Mirror NeuronSystem; ASD groupactivated anterior parietallobe and somatosensorycortex less than thecontrol group in responseto both imitation andaction execution. Leftamygdala activity wasmodulated in the controlgroup by task conditions;activity in this region wassignificantly less variablein the ASD group.



Table 13Group comparisons of regions with significantly elevated likelihood of activation in complex social cognition tasks. Brain areas activated from the ALE analysis (pFDR < 0.05and a minimum cluster size of 200 voxels). C/A: children/adolescents.


x y z

ASD > C; All Superior Temporal Gyrus L 22 536 −48 −32 4 1.59Precentral Gyrus L 4 264 −20 −22 64 1.58Postcentral Gyrus L 3 256 −40 −22 54 1.37Inferior Frontal Gyrus R 13 232 40 22 10 1.36

ASD > C; C/A Superior Temporal Gyrus L 22 472 −48 −32 4 1.58Precentral Gyrus L 4 248 −20 −22 64 1.58Postcentral Gyrus L 3 224 −40 −22 54 1.37Inferior Frontal Gyrus R 13 208 40 22 10 1.36

ASD > C; Adults No clusters found.

C > ASD; All Superior Temporal Gyrus L 41 864 −52 −20 10 1.60Superior Temporal Gyrus R 22 320 56 −8 −2 1.66Inferior Parietal Lobule L 40 208 −54 −56 40 1.52

C > ASD; C/A Superior Frontal Gyrus L 41 832 −52 −20 10 1.60Superior Temporal Gyrus R 22 228 56 −8 −2 1.66Inferior Parietal Lobule L 40 208 −54 −56 40 1.52

tcmtotirrvs1ptrstetiHag2

4

tcihi2caiwtebud

C > ASD; Adults Claustrum R

Examining tasks of visual processing in more detail suggestshey can be considered in three main sub-groups: sensorimotorontrol, visual search and object processing paradigms. Sensori-otor paradigms encompass saccadic and pursuit eye movement

asks (Takarae et al., 2007; Luna et al., 2002) and investigationf antisaccades (Thakkar et al., 2008) and report seemingly con-radictory results with both increased and decreased activationn the cingulate cortex. Studies employing visual search taskseport findings which tend to show less activation in prefrontalegions in ASD compared to control subjects and a greater acti-ation of the occipitotemporal regions in participants with ASD,upporting a more visually based processing strategy (Ring et al.,999; Keehn et al., 2008; Bölte et al., 2008). Object processingaradigms are often carried out as comparison or high level con-rol tasks during other visual or face processing tasks and as suchesults often go unreported. In addition, the choice of task andtimuli varies dramatically between studies, presumably becausehey are chosen to match properties of the main task underxamination. This hetereogeneity is reflected in the results fromhose papers which have reported them, which include areas ofncreased (Keehn et al., 2008) and decreased (Bölte et al., 2008;ubl et al., 2003; Humphreys et al., 2008) activations in individu-ls with ASD, as well as no differences in activation between theroups (Bird et al., 2006; Pierce and Redcay, 2008; Schultz et al.,000).

.2.3. Executive function tasksExecutive dysfunction is well established as an important fea-

ure in ASD (Hill, 2004). When all executive function tasks wereonsidered together in the ALE meta-analysis, greater activation inndividuals with ASD was seen in the orbitofrontal cortex, whichas been associated with learning and planning through the mon-

toring of reward and punishment (Kringelbach and Rolls, 2003,004). Reduced activation was seen in the dorsolateral prefrontalortex, inferior parietal lobe, the insula, posterior cingulate gyrusnd lentiform nucleus. One possible explanation for these findingss that in individuals with ASD, executive dysfunction is associated

ith disruptions to the automatic attentional systems involvinghe prefrontal-striatal network and the inferior parietal lobe (Haist
t al., 2005), as well as the insula (Kana et al., 2007) which haseen associated with executive control of attention. However, exec-tive function is an umbrella term which encompasses a variety ofifferent sub-domains, discussed below.
368 32 −16 16 1.02

Spatial attention has been examined in several studies, withfindings converging on hypoactivation in prefrontal and parietalregions in individuals with ASD relative to controls (Belmonte andYurgelun-Todd, 2003; Haist et al., 2005; Shafritz et al., 2008). Ithas been suggested that reduced parietal activity may result in dif-ficulty with automatically shifting spatial attention in individualswith ASD (Haist et al., 2005). Reductions in prefrontal and parietalactivity in ASD have also been seen in a study of planning (Just et al.,2007).

Cognitive control tasks are mainly associated with hypoactiv-ity of prefrontal cortical brain regions in ASD study participants,primarily the DLPFC (Dichter and Belger, 2008) and the anteriorcingulate cortex (Koshino et al., 2005, 2008; Kana et al., 2007). Therole of the DLPFC in many executive functions including workingmemory is well established in typically developing subjects (Wanget al., 2010; Brazdil et al., 2007), whereas the anterior cingulatecortex is known to be involved in response inhibition. Further evi-dence for anterior cingulate dysfunction comes from Schmitz et al.(2008) who found greater activation in the anterior cingulate cor-tex during a continuous performance task with associated reward.Several authors have chosen to modulate load during cognitivecontrol tasks. Kana et al. (2007) introduced a working memory com-ponent to a response inhibition task and found that, in additionto underactivity in the anterior cingulate cortex, individuals withASD also showed increased premotor cortex activation and reducedsynchrony of the frontal inhibition network with inferior pari-etal regions and the inferior frontal gyrus. In a different approachDichter and Belger (2007) incorporated social stimuli into their taskwhich led to greater levels of hypoactivation being elicited in theASD group than were seen when non-social stimuli were used. Thusit appears that greater differences in neural response are seen astasks become more complex. There is also some suggestion fromthe literature, that as well as underactivating prefrontal regionsduring cognitive control tasks, individuals with ASD show greateractivation in visuo-spatial regions (Belmonte and Yurgelun-Todd,2003; Dichter and Belger, 2007; Gilbert et al., 2008; Koshino et al.,2005; Schmitz et al., 2006) suggesting that they use more visualimagery.

In summary, evidence from a variety of task types suggests
that areas of the prefrontal cortex generally recruited for execu-tive functions are activated to a lesser extent in ASD. These seemto particularly include regions which are involved in the control ofattention.


Table 14Connectivity analysis of fMRI data. Abbreviations as in Table 2.



N (M:F) Mean age (sd) Mean IQ (sd) Diagnosis Diagnosticmeasures

Bird et al.(2006)

14:2 33.3 (11.5) 119 (14) 1 autism, 15AS

DSM-IV,ADOS, AQ

Age, IQ, gender,task performance

1. Localiser task:passive viewing ofneutral faces; passiveviewing of houses;fixation baseline.(fixation cross over‘eye region’)2. Same/differentdiscrimination inhorizontal or verticalplane via buttonpress; pairs of facesand housespresented.

ASD group shows a lackof attentionalmodulation ofV1-parahippocampaland V1-fusiformconnectivity.

Cherkasskyet al. (2006)

53:4 24 (10.6) 106 (16.2) Autism DSM-IV,ADOS

Age, IQ, gender Resting state Compared to controlsASD group showreduced functionalconnectivity in the lefthemisphere generally,between anterior andposterior cingulate,and between the leftparahippocampalregion and all otherROIs

Grèzes et al.(2009)

10:2 26.6 (10.4) 102 (20.6) 10 AS, 2 HFA DSM-IV Age, IQ,behaviouralperformance inscanner

Oddball paradigm;button press toupside down stimuli.Blank screenbaseline.1. Fearful dynamicbody stimuli2. Neutral dynamicbody stimuli3. Fearful static bodystimuli4. Neutral static bodystimuli

Effective connectivityanalysisControl > ASD:amygdala withsuperior temporalsulcus, premotor areaand inferior frontalgyrus; premotor areawith superior temporalsulcus; inferior frontalgyrus with premotorarea

Just et al.(2004)

17 Not reported >80 HFA ADI, ADOS Age, IQ, gender,socioeconomicstatus

Sentencecomprehension(identifying theagent or recipient ofthe action); fixationbaseline.

Control > ASD: leftinferior extrastriatewith left inferiorparietal lobe and leftinferior temporal lobe;left pars triangulariswith superior medialfrontal paracingulate;left inferior temporallobe with left frontaleye field; rightdorsolateral prefrontalcortex left intraparietalsulcus, left inferiorfrontal gyrus, leftinferior parietal lobe,left inferior extrastriateand occipital pole; leftinferior frontal gyruswith calcarine fissure

Just et al.(2006)


Indicate via buttonpress requirednumber of moves tocomplete Tower ofLondon taskdisplayed; fixationbaseline.

Control > ASD:fronto-parietalconnectivity






Just et al.(2007)


Indicate via buttonpress requirednumber of moves tocomplete Tower ofLondon taskdisplayed, Easycondition and hardcondition; fixationbaseline.

Controls > ASD:bilateral inferior andsuperior parietal area,angular gyri, superior,and middle occipitalareas, middle frontalgyri, and the rightprecentral gyrus,superior frontal andthe left inferior frontalgyriAutism > Controls:bilateral hippocampus,thalamus and leftlingual gyrus.Autism > Controls inright middle occipitalgyrus for hard vs easycontrastFunctional connectivityanalysisSee Table 12

Kana et al.(2006)


1. Indicate via buttonpress whether alow-imagerysentence is true orfalse; fixationbaseline.2. Indicate via buttonpress whether ahigh-imagerysentence is true orfalse; fixationbaseline.

ASD group showedgenerally lowerconnectivity betweenlobes than controls butdifferences werenon-significant

Kana et al.(2007)

11:1 26.8 (7.7) 110.1 (12.6) HFA ADI, ADOS Age, IQ, gender,handedness,socioeconomicstatus

Go/no go task withincreasing workingmemory load: 1 –simple responseinhibition, 2 –response inhibitionwith workingmemory component;fixation baseline.

Control > ASD:Inhibition network(cingulate cortex,cingulate gyri andinsula) with rightinferior parietal areasand right inferiorfrontal gyrus

Kennedy andCourchesne(2008)

12:0 26.5 (12.8) 101.6 (15.2) 6 autism, 6AS

ADI-R, ADOS Age Counting Stroop taskwith emotional,neutral and numberwords. Participantsasked to countnumber of words onthe screen. Baselinefixation cross

Functional connectivityanalysis of taskpositive network andtask negative networkASD individualsshowed reducedconnectivity within thetask negative network,localised to the medialprefrontal cortex andthe left angular gyrusNo between groupdifferences withrespect to task positivenetwork

Kleinhans et al.(2008b)

19:0 23.5 (7.8) 106.7 (15.7) 8 autism, 9AS, 2PDD-NOS

DSM-IV,ADI-R, ADOS


1-back with neutralfaces; 1-back withhouses as baseline.

Control > ASD: rightfusiform face area withleft amygdala, bilateralposterior cingulate andleft cuneusSocial impairmentcorrelated with reducedconnectivity betweenright fusiform face areaand left amygdala, andwith increasedconnectivity betweenright fusiform face areaand left inferior frontalgyrus






Koshino et al.(2005)

13:1 25.7 100.1 HFA ADI, ADOS Age, IQ, gender,socioeconomicstatus, taskperformance

n-back task (0,1,2);fixation baseline.

Control > ASD: leftinferior parietal lobewith right dorsolateralprefrontal cortex, rightfrontal eye fields, rightposterior precentralsulcus, leftintraparietal sulcus,right intraparietalsulcus and rightsuperior parietal lobe;left intraparietal sulcuswith superior medialfrontal paracingulate

Koshino et al.(2008)



Control > ASD: leftfrontal-fusiformconnectivity

Lee et al. (2009) 9:3 10.2 (1.6) 113.3 (17.3) ASD DSM-IV,ADI-R, ADOS

Age, gender. IQ Go/NoGo task Functional connectivityanalysis of inferiorfrontal gyrusNo group differencesbut ASD group showeda negative relationshipbetween age and rightinferior frontal gyrusconnectivity withbilateralsupplementary motorarea and right caudatewhich was not seen incontrols

Mason et al.(2008)

17:1 26.5 101.9 HFA ADI, ADOS Age, IQ, gender,socioeconomicstatus, race

Read short scenariosand answered yes/nocomprehensionquestions via buttonpress.Intentional,emotional andphysical sentences;fixation baseline.

Control > ASD: Leftmedial frontal cortexwith lefttemporal-parietaljunction (emotional);left medial frontalcortex with left inferiorfrontal and middletemporal gyri and righttemporo-parietaljunction (intentional)

Mizuno et al.(2006)

8:0 28.4 (8.9) 86.5 (11.4) Autism DSM-IV,ADI-R, CARS


Image of hand withdot indicatingappropriate buttonpress for a 6-digitrepeated sequence;single-digit stimuli asbaseline.

Functional connectivityanalysis of thalamus:Control > ASD:Right thalamus withbilateral superiorfrontal gyri, leftparacentral gyrus, rightprecuneus, left medialtemporal lobe, rightsuperior temporal lobeand rightparahippocampal gyrusLeft thalamus withbilateralparahippocampal gyrusand right superiortemporal gyrusASD > Control:Right and left thalamuswith bilateral insula,middle frontal gyrus,medial frontal region,precentral andpostcentral gyri andleft inferior parietallobe. Left thalamuswith left cuneus andsuperior temporalgyrus






Noonan et al.(2009)

10:0 23.0 (9.9) 96.7 (16.1) 6 autism, 4AS

DSM-IV,ADI-R, ADOS

Age, gender, IQ Source memory task:Participants givenlists of words toremember in visualand auditorydomains thenpresented with morelists of words andasked to judgewhether they werenew, previouslypresented in thevisual domain orpreviously presentedin the auditorydomain; blank screenbaseline.Analysis performedonly on visualrecognition runs

ROIs chosen becausethey showed distinctpatterns of activationbetween groupsASD > Control:Left middle frontalgyrus with bilateralprecentral gyrus andsupplementary motorarea, bilateral superiortemporal, left middletemporal, rightparahippocampal, leftsupramarginal and leftpostcentral gyri andleft precuneusLeft superior parietallobe with bilateralsupplementary motorarea and left precentralgyrus, right inferiorfrontal, bilateralsuperior and middletemporal gyri, leftsuperior and inferiorparietal lobules, leftfusiform and leftlingual gyri and leftcerebellumLeft middle occipitalgyrus with bilateralinferior frontal andsuperior temporal gyri

Turner et al.(2006)




Functional connectivityanalysis of caudatenucleus:Control > ASDBilateral caudate withright superior frontalgyrus. Right caudatewith left middletemporal gyrus, rightparahippocampalgyrus and bilateraloccipital cortexASD > ControlBilateral caudate withright middle frontalgryus, bilateralprecentral gyrus, leftmedial frontal gyrus,right postcentral gyrus,bilateral cingulategyrus, and left cuneus.

Villalobos et al.(2005)




Functional connectivityanalysis of visualcortex:Control > ASD:Bilateral inferiorfrontal area, rightsuperior frontal gyrusand paracentral lobule,bilateral thalamus,right basal ganglia andcerebellar vermis






Welchew et al.(2005)

13:0 31.2 (9.1) 108.6 (17.1) Aspergersyndrome

DSM-IV, AQ age, IQ, gender,handedness

Implicit emotionalface processingparadigm.Participants viewedfearful faces,non-fearful faces,neutral faces andscrambled faces.Asked to press abutton whenstimulus appeared

Reduced functionaldistance betweenmedial temporalstructures (bilateralparahippocampal gyrusand left amygdala) andrest of brain

Wicker et al.(2008)

11:1 27 (11) 81.3 8 autism, 4AS

DSM-IV Age Explicit angry –happy face emotionjudgements; baselineexplicit genderjudgement. Actorsgaze either direct oraverted

Effective connectivityanalysisControl > ASDDorsomedial prefrontalcortex withdorsolateral prefrontalcortex, occipital cortexwith fusiform gyrus,dorsolateral prefrontalcortex withventrolateralprefrontal cortexASD > ControlDorsolateral prefrontal

4

tiwg

irgi2irgcafbecAceiirtc

ooDaTar

.2.4. Auditory and language tasksGiven the relative primacy of communication dysfunction in

he diagnostic triad, there have been surprisingly few fMRI stud-es which have directly examined language function in individuals

ith ASD, although a much greater number of studies rely on lan-uage processing for successful task completion.

The ALE meta-analysis revealed clusters of reduced activationn individuals with ASD in the bilateral superior temporal gyri, aegion which is well known to be associated with receptive lan-uage. Reduced activation in this region in individuals with ASDn response to spoken language (Gervais et al., 2004; Wang et al.,007) may underlie some of the verbal communication difficulties

n ASD. Less activation was also seen in individuals with ASD in theight pyramis of the cerebellar vermis and the left middle cingulateyrus. Increased activations were found in individuals with ASDompared to controls in the motor cortex, the cerebellar declivend the posterior cingulate. The role of the cerebellum in languageunction is increasingly recognised (Stoodley, 2011) and the cere-ellar declive has been previously associated with dyslexia (Pernett al., 2009). It is not necessarily clear why the motor cortex or theingulate cortex should be activated differently in individuals withSD during language tasks. Some studies have identified motorortex activation during language tasks (Hauk et al., 2004; Floelt al., 2003) and it is possible that this occurs to a greater degree inndividuals with ASD due to the use of atypical language process-ng strategies. Alternatively it may be that the differences are taskelated with studies involving forced choice and target detectionasks differentially recruiting the premotor and anterior cingulateortices in ASD.

In keeping with the idea that language processing in ASD occursutwith ‘traditional’ language regions, reductions in the lateralityf language regions have been identified in a number of studies.uring a response naming task Knaus et al. (2008) identified greater
ctivations in frontal and temporal regions in individuals with ASD.hese included Broca’s area which was also significantly less later-lised in individuals with ASD. Takeuchi et al. (2004) also foundeduced left lateralisation of language functions due to an increase
cortex with fusiformgyrus

in right sided activation during a reading task. Similarly Kleinhanset al. (2008a) showed that individuals with ASD had reduced leftlaterality in a verbal fluency task. These findings are consistent witha study using structural MRI by De Fossé et al. (2004) which foundreversal of asymmetry in frontal language regions in individualswith autism.

4.2.5. Basic social processingAs social deficits are a core feature of ASD and much social infor-

mation during everyday communication is conveyed within faces,it is unsurprising that face processing has attracted a lot of atten-tion in the fMRI investigation of ASD (Williams, 2006; Itier andBatty, 2009). The main body of fMRI research in ASD concentrateson the investigation of networks underlying face perception andinterpretation of facial expressions and emotions.

The processing of faces in individuals with ASD appears to beatypically organized in comparison to the network seen in con-trol subjects. Activations in the fusiform face area (FFA – Haxbyet al., 2002), and the occipital face area are commonly reportedas reduced in individuals with ASD (Humphreys et al., 2008;Hadjikhani et al., 2007; Pierce et al., 2001; Bookheimer et al., 2008;Kleinhans et al., 2008a,b) and this is reflected in the results of ourmeta-analysis. In the meta-analysis by Di Martino et al. (2009) ofsocial tasks in ASD, a relative hypoactivation was also reported inthe left fusiform and bilateral occipital lobe.

Schultz et al. (2000) hypothesized that ASD participants lackingexpertise in processing faces so that they would therefore processface stimuli in a manner similar to object processing which is “typ-ically reliant on the detection of individual features” and thereforeless susceptible to the inversion effect of faces. Their fMRI findingssupport the hypothesis that in ASD the neural network recruited toprocess faces is more akin to that used for object processing. Pierceet al. (2001) confirmed FFA hypoactivation in ASD and also found
that individuals with ASD showed maximal activation during faceprocessing in a wide variety of other areas, including the cerebel-lum and frontal lobe. In combination with other studies suggestingthat the FFA may respond not only to faces but also to stimuli for

9 Biobe

wtuf

rriffttvDiw

rtgtept2

aoasbthug(2pn

teb(CdaTcmsdetwvWvbnte

ias

36 R.C.M. Philip et al. / Neuroscience and

hich specialization has occurred in response to experiential fac-ors (Gauthier et al., 2000), this suggests that individuals with ASDse an object processing style due to a lack of developed expertiseor faces.

Differences in eye scan path when examining faces has beeneported in the behavioural literature leading to the suggestion thateports of hypoactivation in typical face processing brain regionsn ASD result from ASD participants processing faces using a dif-erent method rather than being reflective of a difference in theunctional neuroanatomy of people with autism. In keeping withhis idea, studies which have directed participants eye gaze towardshe eye region have reported a lack of difference in fusiform acti-ation (Hadjikhani et al., 2004a, 2007; Bird et al., 2006). Similarlyalton et al. (2005) reported hypoactivation of the fusiform gyrus

n individuals with ASD when viewing faces but this was associatedith spending less time fixating on the eyes of the faces.

We also found aberrant activity in the superior temporal sulcusegion in individuals with ASD compared to controls (increased inhe superior temporal gyrus and reduced in the middle temporalyrus) in line with the STG over and under-activation reported inhe meta-analysis by Di Martino et al. (2009). In addition to the wellstablished role of this superior temporal sulcus region in auditoryerception, increasing evidence also points to an important role ofhis region in the interpretation of social stimuli (Zilbovicius et al.,006).

Although it was not identified in our meta-analysis, aberrantctivation of the amygdala has been reported in a number of studiesf face processing. Dalton et al. (2005) reported amygdala hyper-ctivation which was related to time spent fixating on the eyes anduggested this may be because individuals with ASD find eye gaze toe an anxiety provoking stimulus. Kleinhans et al. (2009) reportedhat repeated exposure to face stimuli led to apparent amygdalayperactivation in the ASD group due to a lack of amygdala habit-ation to neutral faces typically seen in controls. In contrast, otherroups report reduced amygdala activation in individuals with ASDHadjikhani et al., 2007; Bookheimer et al., 2008; Corbett et al.,009). Differences in the stimuli used and instructions given mayossibly account for these inconsistencies e.g. emotional versuson-emotional faces, passive viewing versus active response.

Studies of emotional faces are complicated by the joint inves-igation of face processing and emotional processing in the samexperiment. Several studies have reported modulation of socialrain areas in response to the intensity of emotion presentedAshwin et al., 2007; Critchley et al., 2000; Deeley et al., 2007).ritchley et al. (2000) also manipulated task demands to exploreifferences between groups if the emotional content was explicitlyttended to by participants (emotion label) or not (gender label).his resulted in a significant group by condition interaction in theerebellum, striatum, insula, amygdalohippocampal junction andiddle temporal gyrus. Further evidence of the importance of the

timuli presented comes from Pelphrey et al. (2007) who found noifference between individuals with ASD and controls when staticmotion stimuli were used but greater activation of the superioremporal sulcus region in the ASD group when dynamic stimuliere employed. This led the authors to suggest that aberrant acti-

ation of social brain regions in ASD is specific to dynamic emotion.hilst this is in contrast with some previous reports of hypoacti-

ation in ASD cohorts in response to static displays of faces, it maye that in this study group differences were sub-threshold witheutral and static face stimuli and the effect is more robust usinghese more salient and environmentally valid stimuli like dynamicmotion morphs.

The two fMRI studies which analysed effective connectivityn face-selective networks in ASD present advanced explanatorypproaches of the differently activated networks in ASD and controlubjects. For a mechanistic understanding of a network, effective

havioral Reviews 36 (2012) 901–942

connectivity measures provide causal influences between neuronsand neuronal population, which can be interpreted as connec-tion strengths between regional neuronal populations within anetwork. Both effective connectivity studies provide evidence forsignificant alterations in face-selective processing in ASD partici-pants (Wicker et al., 2008; Bird et al., 2006). Connection strengthsfor facial emotion processing are greater in control groups in twomain aspects: firstly, controls show a greater connection strengthbetween the V1 and the FFA than individuals with ASD. Secondly,prefrontal brain regions, the amygdala and the STS are involved inthe network to a greater extent in control individuals than thosewith ASD (Wicker et al., 2008; Bird et al., 2006). Not only do bothstudies on effective connectivity support findings on reduced acti-vations in face-selective regions but they also suggest that thisis modulated by underconnectivity with prefrontal and occipitalregions.

There is emerging behavioural evidence that social deficits inASD may not be limited to facial stimuli or indeed the visual domain(Philip et al., 2010). Several studies have also been designed tospecifically address the processing of movement from social stim-uli i.e. biological motion (Freitag et al., 2008; Grèzes et al., 2009;Herrington et al., 2007). These provide evidence that in ASD, thebrain processes movement in a different way but particularly whenthe movement conveys information that is socially relevant. Todetermine the effect of emotional content in biological motionGrèzes et al. (2009) conducted a study examining brain activationin individuals with ASD in response to fearful and neutral emo-tional body movements. They found that when the BOLD responseto fearful gestures was compared to that for neutral gestures indi-viduals with ASD showed hypoactivation of the right inferior frontalgyrus, precentral gyrus, inferior temporal gyrus and the amygdala;they also showed overactivation of the left medial anterior superiorfrontal gyrus. The authors suggest that whilst the systems whichunderlie the detection and representation of action are intact inpeople with ASD, the brain regions involved in emotion processingare not and that this relates to reduced connectivity between theamygdala and prefrontal cortex.

4.2.6. Complex social cognition tasksSocial cognition deficits in ASD are most frequently discussed

in terms of a mentalizing deficit or Theory of Mind dysfunction(Baron-Cohen et al., 1999). The theory states that individuals withASD have difficulty attributing mental states to others and this maystem from difficulties in processing social information and cuesfrom others. These cues may be in the form of facial expression butalso extend to body posture, tone of voice and the more generalcontext of the situation.

The meta-analysis of complex social cognition tasks revealedevidence of aberrant activation in the superior temporal gyri (bothincreased and decreased activations were apparent in people withASD). As discussed above, there is increasing evidence for the roleof superior temporal sulcal regions in social cognition and our find-ings and others (Di Martino et al., 2009) suggest that this region isaffected in ASD.

Aberrant activations were also seen in the right inferior frontalgyrus (increased in ASD) and left inferior parietal lobule (decreasedin ASD). Both of these are regions have been previously shown toactivate differently within social tasks in ASD (Di Martino et al.,2009) and are thought to form part of the mirror neuron sys-tem (MNS) in humans. The MNS comprises populations of neuronswhich have been shown in animal studies to activate to both theexecution of a particular motor action and in response to observing
that same motor act performed by another (Gallese et al., 1996).It has been proposed that for successful complex mentalizing totake place, another’s cognitive perspective must be assessed andre-represented following the observation of their actions (Uddin

Biobe

eeicrveiuwn

obTvodfe

4

4

fi(frinctufat(iiaftf(ti

4

igapfdvsa

tvdtg


t al., 2008). The MNS has thus been proposed as a system enablingmotion understanding via action representation. On a neural levelt is thought that whilst the inferior frontal gyrus and parietal cortexode the observed action (Freitag et al., 2008), in the case of sociallyelevant and emotional stimuli the limbic system is also recruitedia the insula, allowing internally felt emotional significance (Carrt al., 2003; Schulte-Rüther et al., 2007). By understanding and cod-ng the function of others actions, intention can be attributed and annderstanding of another’s mental state can be achieved. Togetherith these studies our meta-analysis provides evidence for mirroreuron dysfunction in ASD.

In neither our analysis of basic or complex social tasks did webserve aberrant activity of the anterior cingulate cortex, reportedy Di Martino et al. (2009) in their meta-analysis of social tasks.his discrepancy may be accounted for by differences in subdi-ision of tasks as well as differences in approach taken to carryut the meta-analysis; Di Martino et al. extracted within groupata and contrasted this whereas we extracted significant resultsrom between group contrasts and calculated activation likelihoodstimations from this.

.3. Emerging themes across task domains

.3.1. The influence of specific task/stimuli demandsNot only the stimulus type but also the task demands are critical

actors when designing tasks to elucidate brain function differencesn ASD. Whilst some studies suggest task independent dysfunctionKennedy and Courchesne, 2008), many apparent contradictions inMRI data using tasks and stimuli of a similar nature with compa-able study populations could be accounted for by the specific tasknstructions given. For example, Critchley et al. (2000) report a sig-ificant group x task interaction suggesting that the ASD group andontrol group responded differently to whether the task requiredhem to explicitly attend to the emotional content of face stim-li or not. Differences in fusiform gyrus activity in response toace processing has been reported when participants carried out

perceptual emotion match task, but not a more linguistic emo-ion label task even though both tasks employed the same stimuliPiggot et al., 2004; Wang et al., 2004). Further, when investigat-ng the effect of face familiarity on brain activation in autism, anmplicit task where the condition of familiarity did not have to bettended to for successful task completion failed to reveal any dif-erences between groups (Pierce et al., 2004); whereas when theask employed required participants to discriminate on the basis ofamiliarity, fusiform hypoactivation was reported in the ASD groupDalton et al., 2005). It is therefore important to carefully considerhe question to be answered when designing tasks and also whennterpreting data.

.3.2. Lack of preference for social stimuliAlthough we identified fusiform gyrus hypoactivation in ASD

n the meta-analysis, there are multiple reports of typical fusiformyrus function in autism. In studies which show typical fusiformctivation tasks have been designed to incorporate cues to ensurearticipants engage with the face stimuli and attend to the salienteatures of faces such as eyes. This suggests that face processingifferences in individuals with ASD may result from a lack of moti-ation to attend to, or a preference for not attending to, the sociallyalient features of faces, as opposed to a deficit in the fusiform facerea per se.

Interestingly, there are several examples in the literature ofasks which identify no between group differences in brain acti-
ation, yet when social stimuli are incorporated into the task aeficit in the ASD group becomes evident. For instance, when inves-igating brain responses to congruence, hypoactivation in the ASDroup was not found in the task employing arrow stimuli but only
havioral Reviews 36 (2012) 901–942 937

was when face stimuli were used (Dichter and Belger, 2007). Simi-larly, when attention modulation was investigated, brain responseswhen house stimuli were used did not differ between the ASD groupand the control group. However, the same task involving face stim-uli revealed a hypoactivation in the ASD group (Bird et al., 2006). Inthe auditory domain this pattern has been repeated. No differenceswere reported in response to environmental sounds, however vocalsounds resulted in less activation in the ASD group relative to thecontrol group, with 3 of the 5 individuals with autism tested show-ing no superior temporal sulcus activation (Gervais et al., 2004).Both tasks that investigated biological motion reported hypoac-tivation in response to biological motion in the ASD group thatwas either not present or far less extensive than in response tomotion derived from a non-biological source (Freitag et al., 2008;Herrington et al., 2007). These studies suggest that typically devel-oping individuals show an increase in brain activation in responseto social stimuli, possibly due to increased attention to them, whichis not seen in ASD.

In each of the aforementioned studies, although participantswere asked to attend to the stimuli, they were not required to makea social judgement based upon them. Studies which have incor-porated explicit social judgements, in contrast to the above, tendto show hyperactivation of the social brain network (Baron-Cohenet al., 1999; Wang et al., 2006), perhaps reflecting less efficient, orincreased effortful, processing.

4.3.3. Lack of modulation in response to task/stimuli demandsAnother repeated finding in tasks of various types, using a range

of stimuli, is that whilst between group contrasts fail to identify anydifferences between ASD and control samples, suggesting that thetypical neural circuitry is recruited in both groups, subtle modula-tion of this activation in response to task demands or intensity ofstimuli is observed in the control group but not in the ASD group.Whilst the appropriate brain network is crudely recruited, finer andmore subtle control of activity in these brain regions is lacking.

In tasks of facial emotion processing where the intensity of stim-uli modulated brain responses in the control group, this additionallevel of responsiveness to the stimuli was not observed in the ASDgroup (Ashwin et al., 2007; Deeley et al., 2007). In another faceprocessing task, this time of neutral stimuli, whilst there were nodifferences between groups in response to task stimuli the controlgroup modulated brain activity in response to attention demands,an effect not seen in the ASD group (Bird et al., 2006). Daltonet al. (2005) reported a modulatory effect on the fusiform gyruswith regard to the familiarity of the face stimuli presented in thecontrol group but this effect was not present in the ASD group.Motion has been shown to modulate facial emotion processing incontrols but not participants with ASD (Pelphrey et al., 2007). Inboth studies investigating the neural response to the congruenceof stimuli, similar circuitry was recruited by both the ASD and con-trol groups in response to processing information from eye gaze.However whether eye gaze was congruent or not with the appear-ance of a target, modulated the activity in this system in the controlgroup but not the ASD group (Dichter and Belger, 2007; Pelphreyet al., 2005). A lack of modulation in ASD has also been reportedin lexical tasks. Again, broadly, the control and ASD participantswere reported to recruit typical neural circuitry for the task butthe control group modulated this in accordance with the type ofword (Harris et al., 2006) or sentence (Mason et al., 2008) pre-sented whereas the ASD group did not. During an imitation task,specific task conditions were found to affect the levels of activityin the mirror neuron system and amygdala in the control group
and whilst these brain regions did activate in the ASD group thevariability in response to conditions was not observed (Williamset al., 2006). Also in tasks involving understanding the intentionsof others, whilst control group activation varied in relation to the

9 Biobe

jcf

4

sewItttpwbitthdostto

4

mtd2ecst

nscnstg(e

4

tiviyrtct

tfigoui


udgement they were making (Pinkham et al., 2008) or the ironicontent of stimuli (Wang et al., 2007), these features of the taskailed to modulate brain activation in the ASD group.

.3.4. Visuo-spatial processing styleCumulatively, there are findings from a variety of task types to

uggest that areas of the prefrontal cortex generally recruited forxecutive functions are activated to a lesser extent in ASD. Moreeight is however generally given to visual and perceptual systems.

t is not obvious how a visually predisposed cognitive style relateso the core clinical features of autism but it is interesting to note thathe evidence based and most commonly adopted approaches usedo teach children with autism rely heavily on the use of visual sup-orts. In establishing basic requesting skills in non-verbal childrenith autism, the Picture Exchange Communication Systems has

een found to be effective (Bondy and Frost, 2001) and is commonlymplemented by Speech and Language Therapists in schools acrosshe UK. In classrooms, the TEACCH approach (Treatment and Educa-ion of Autistic and related Communication handicapped Children)as been found to be an effective method for working with chil-ren with autism (Probst and Leppert, 2008) and relies on a rangef visual measures (visual timetables, start/finish boxes, timers) toupport children with autism through their school day. It may behat the effectiveness of these strategies at least in part is due tohe fact that they capitalize on the naturally visual cognitive stylef people with autism.

.3.5. DysconnectivityIn recent years increasing interest has focused on the idea that

any of the deficits in ASD result from changes to brain connec-ivity. Evidence from electroencephalography (Hughes, 2007) andiffusion tensor imaging (Barnea-Goraly et al., 2004; Lee et al.,007a) as well as structural MRI studies of white matter (Herbertt al., 2003) all support this idea. The majority of fMRI studies haveoncentrated on single brain regions which, whilst interesting, isomewhat artificial and does not take into account the influencehat distributed networks may have on brain activity.

Findings from the review suggest that reductions in brain con-ectivity are widespread and not task specific. Some studies haveuggested that decreases are primarily confined to cortico-corticalonnectivity with increases occurring in subcortico-cortical con-ections (Mizuno et al., 2006; Turner et al., 2006). This increase inubcortico-cortical connectivity might be explained by compensa-ion for reduced cortico-cortical connectivity or compensation forreater arousal modulation in ASD subjects than in healthy subjectsBelmonte et al., 2004). Future research should hopefully clarify thexact nature and distribution of connectivity differences in ASD.

.3.6. Age-related changesThe age group comparisons revealed different activated clus-

ers in the different age groups compared with the clusters foundn the whole-group analysis. Both age groups showed different acti-ated clusters within the same network and different activationsn different networks. In addition, the individual age group anal-ses compared with the main analyses in the same task domainsevealed substantial differences in the amount of significant clus-ers in the networks, the probability of activations in the samelusters and the probability of activations in different clusters inhe network.

The age-separated meta-analyses in auditory and languageasks, social tasks and complex cognition tasks reveal differentndings in child and adolescent populations than in adults, sug-
esting that different developmental changes in brain activationccur in individuals with ASD than in typically developing individ-als. Different patterns of brain development have been reported
n structural MRI studies of ASD (Stanfield et al., 2008; Redcay and


Courchesne, 2005); our findings suggest that this may also be thecase for brain function. This may relate to innate neurobiologi-cal differences which affect brain development but may also besecondary to experiential factors and learnt compensatory mech-anisms in individuals with ASD. One caveat to our findings is thatthe separate meta-analyses conducted for the different age groupsrelied upon small numbers of studies and at present these findingsmust be regarded as preliminary in nature.

4.4. Limitations of the current review

The results of the meta-analysis will obviously depend uponwhich tasks are allocated to which domain and previously stud-ies have taken a different approach to ours (Di Martino et al., 2009).Combining dissimilar tasks risks increasing heterogeneity, whereasoverenthusiastic subdivision leads to a lack of power. We havetried to strike a balance between these opposing strategies butacknowledge the somewhat arbitrary decisions that this entails.For example, many of the tasks which we considered to be visualin nature also contained components of executive function. This isparticularly true for visual search or embedded figures tasks wherethe consideration of context is paramount. It is not clear at whatpoint a task becomes more visual and less executive in nature orvice versa and this is further complicated by the fact that individu-als with ASD seem to recruit visual regions more than controls doto accomplish executive tasks. Decisions regarding the sub-divisionof social tasks into simple and complex were also difficult to makebut we feel that the distinct areas of activation identified for each ofthese domains to some extent validates this approach. Participantsappeared to overlap between studies, although different tasks wereused in each. Thus findings from the meta-analyses may be moreheavily weighted towards differences in the populations that havebeen published on most frequently, particularly if they examinedtasks from the same domain. Finally, we excluded studies whichwere not published in English which may bias the results some-what.

5. Future directions

It is clear that for pragmatic reasons the majority of the ASDcohorts studied using fMRI are not entirely representative of theASD population as a whole. Whilst findings from these samples doadd a valuable contribution to the understanding of autistic neu-rophysiology the interpretation of findings should be extremelycautious. Effort is therefore required to address the sample selec-tion issues in this field and careful design of simple tasks may allowthe study of lower IQ and younger cohorts. The use of simple tasksmay also shed light on more basic cognitive processes which mayunderlie some of the more complex deficits seen in ASD. Greaterinvestigation of children and adolescents will allow developmen-tal changes in brain function to be examined and also shed light onwhether differences identified in adult populations are related tothe expression of ASD or represent mechanisms developed to com-pensate for impairment. High-risk studies of very young infants,although ethically challenging, would provide vital informationregarding brain development prior to the onset of observablefeatures of ASD and may facilitate the development of predictivetools and preventative interventions.

The small sample size investigated in many studies along withthe use of liberal statistical thresholds also make it difficult todraw firm conclusions from the research to date. It is increasingly
acknowledged that ASD, or even autism itself, does not representa single disorder but is likely to be the result of multiple under-lying causes leading to similar clinical features – the ‘autisms’(Baumann, 2010; Geschwind and Levitt, 2007). It is not surprising

Biobe

teboevtestllpbi

btaoasai

dfbbwtadua

rtugaciuetiie

6

taItt

A

fHR


herefore that heterogeneity exists in the current neuroimaging lit-rature. To reduce this, future studies could examine subgroupsased on clinical phenotypes; this would maximize the powerf studies to identify significant differences between groups andxamine the effect of putative aetiological mechanisms. The con-erse approach may also be useful, i.e. using functional imagingo derive subgroups based upon patterns of activation and thenxamining the clinical features associated with these. Subdividingamples on the basis of putative genetic or environmental risk fac-ors and examining the brain activation associated with these is alsoikely to be a fruitful approach. All of these approaches will requirearge samples and detailed clinical characterisation of researcharticipants. The progression of multicentre imaging as proposedy Belmonte et al. (2008) will allow for larger samples to be

nvestigated.Using meta-analysis, as we have, to look for replicated findings

etween studies is one way of attempting to overcome the limita-ions inherent in the literature due to small sample sizes. However,s we have discussed, heterogeneity may also result from the usef even subtly different stimuli or task designs between studies,dding further statistical noise to the meta-analysis. Replicationtudies are given relatively low priority by investigators, fundersnd journal editors alike; we would politely suggest that this policys revisited.

Task design in general, including the selection of baseline con-itions, requires careful consideration as both task demands andeatures of stimuli appear to have differential effects in ASD. Theroad range of tasks used to investigate ASD which report atypicalrain activation also require investigators to be aware that evenhen designing a task with a specific element of interest, assump-

ions made about other aspects of the task may not be the sames those that could be made about a typical population. Parametricesigns may be particularly informative when studying this pop-lation due to the consistent reports of aberrant modulation ofctivation in ASD.

There are a number of more specific areas identified by thiseview that would benefit from greater investigation. It is clearhat there has been an emphasis on the processing of social stim-li in the ASD literature, in particular faces. This is not surprisingiven that social communication difficulties form a core basis of theutistic phenotype. However, impairment is not limited to face pro-essing or the visual domain; investigating the neural response ofndividuals with ASD when processing other kinds of social stim-li, such as voice tone and movement, particularly in relation tomotion may illuminate regions of the brain which are commono socio-emotional processing, regardless of the form in whicht is presented. Other areas which appear to have been under-nvestigated to date include communication, motor function and,specially, sensory hypersensitivity.

. Conclusion

Despite the difficulties identified during this review it is clearhat fMRI has added greatly to our understanding of ASD and therere a number of consistent themes running through the literature.mprovements in paradigm design, data acquisition and analysisechnology are likely to further refine our understanding of ASD inhe future.

cknowledgments

The authors report no conflicts of interest. RCMP and KB wereunded by a Medical Research Scotland grant awarded to ACS.CW is supported by a Dorothy Hodgkin Fellowship from theoyal Society. During this work ACS was supported by a Wellcome


Trust Clinical Research Training Fellowship. Further support for thisstudy came from the Patrick Wild Centre.

References

Allen, G., Muller, R.-A., Courchesne, E., 2004. Cerebellar function in autism: functionalmagnetic resonance image activation during a simple motor task. BiologicalPsychiatry 56, 269–278.

Allen, G., Courchesne, E., 2003. Differential effects of developmental cerebellarabnormality on cognitive and motor functions in the cerebellum: an fMRI studyof autism. American Journal of Psychiatry 160, 262–273.

American Psychiatric Association, 2000. Diagnostic and Statistical Manual of Men-tal Disorders, 4th ed. text rev; DSM-IV-TR. American Psychiatric Association,Washington, DC.

Ashwin, C., Baron-Cohen, S., Wheelwright, S., O’Riordan, M., Bullmore, E.T., 2007.Differential activation of the amygdala and the ‘social brain’ during fearful face-processing in Asperger Syndrome. Neuropsychologia 45, 2–14.

Asperger, H., 1944. Die ‘Autistischen Psychopathen’ im Kindesalter. In: Frith, U.(1991). Autism and Asperger syndrome. ‘Autistic psychopathy’ in childhood.Cambridge University Press, pp. 37–92.

Barnea-Goraly, N., Kwon, H., Menon, V., Eliez, S., Lotspeich, L., Reiss, A.L., 2004. Whitematter structure in autism: preliminary evidence from diffusion tensor imaging.Biological Psychiatry 55, 323–326.

Baron-Cohen, S., Ring, H.A., Wheelwright, S., Bullmore, E.T., Brammer, M.J., Simmons,A., Williams, S.C.R., 1999. Social intelligence in the normal and autistic brain: anfMRI study. European Journal of Neuroscience 11, 1891–1898.

Baumann, M.L., 2010. Medical comorbidities in autism: challenges to diagnosis andtreatment. Neurotherapeutics 7 (3), 320–327.

Belmonte, M.K., Mazziotta, J.C., Minshew, N.J., Evans, A.C., Courchesne, E., Dager, S.R.,Bookheimer, S.Y., Aylward, E.H., Amaral, D.G., Cantor, R.M., Chugani, D.C., Dale,A.M., Davatzikos, C., Gerig, G., Herbert, M.R., Lainhart, J.E., Murphy, D.G., Piven,J., Reiss, A.L., Schultz, R.T., Zeffiro, T.A., Levi-Pearl, S., Lajonchere, C., Colamarino,S.A., 2008. Offering to share: how to put heads together in autism neuroimaging.Journal of Autism and Developmental Disorders 38 (1), 2–23.

Belmonte, M.K., Cook Jr., E.H., Anderson, G.M., Rubenstein, J.L., Greenough, W.T.,Beckel-Mitchener, A., Courchesne, E., Boulanger, L.M., Powell, S.B., Levitt, P.R.,Perry, E.K., Jiang, Y.H., DeLorey, T.M., Tierney, E., 2004. Autism as a disorder ofneural information processing: directions for research and targets for therapy.Molecular Psychiatry 9, 646–663.

Belmonte, M.K., Yurgelun-Todd, D.A., 2003. Functional anatomy of impaired selec-tive attention and compensatory processing in autism. Cognitive Brain Research17, 651–664.

Bird, G., Catmur, C., Silani, G., Frith, C., Frith, U., 2006. Attention does not modulateneural responses to social stimuli in autism spectrum disorders. Neuroimage 31,1614–1624.

Bölte, S., Hubl, D., Dierks, T., Holtmann, M., Poustka, F., 2008. An fMRI-study of locallyoriented perception in autism: altered early visual processing of the block designtest. Journal of Neural Transmission 115, 545–552.

Bölte, S., H.D., Feineis-Matthews, S., Prvulovic, D., Dierks, T., Poustka, F., 2006.Facial affect recognition training in autism: can we animate the fusiform gyrus?Behavioural Neuroscience 120, 211–216.

Bondy, A., Frost, L., 2001. The picture exchange communication system. BehaviourModification 25, 725–744.

Bookheimer, S.Y., Wang, A.T., Scott, A., Sigman, M., Dapretto, M., 2008. Frontal contri-butions to face processing differences in autism: evidence from fMRI of invertedface processing. Journal of the International Neuropsychological Society 14,922–932.

Brazdil, M., Mikl, M., Marecek, R., Krupa, P., Rektor, I., 2007. Effective connectivity intarget stimulus processing: a dynamic causal modeling study of visual oddballtask. Neuroimage 35 (2), 827–835.

Carmody, D., Moreno, R., Mars, A., Seshadri, K., Lambert, G., Lewis, M., 2007. Briefreport: brain activation to social words in a sedated child with autism. Journalof Autism and Developmental Disorders 37, 1381–1385.

Carr, L., Iacoboni, M., Dubeau, M.C., Mazziotta, J.C., Lenzi, G.L., 2003. Neural mecha-nisms of empathy in humans: a relay from neural systems for imitation to limbicareas. Proceedings of the National Academy of Science of the United States ofAmerica 100 (9), 5497–5502.

Cherkassky, V., Kana, R., Keller, T., Just, M., 2006. Functional connectivity in a baselineresting-state network in autism. Neuroreport 17, 1687–1690.

Chiu, P.H., Kayali, M.A., Kishida, K.T., Tomlin, D., Klinger, L.G., Klinger, M.R., Mon-tague, P.R., 2008. Self responses along cingulate cortex reveal quantitative neuralphenotype for high-functioning autism. Neuron 57, 463–473.

Cohen, M.S., Bookheimer, S.Y., 1994. Localization of brain function using magneticresonance imaging. Trends in Neurosciences 17, 268–277.

Constantino, J.N., Davis, S.A., Todd, R.D., Schindler, M.K., Gross, M.M., Brophy, S.L.,Metzger, L.M., Shoushtari, C.S., Splinter, R., Reich, W., 2003. Validation of a briefquantitative measure of autistic traits: comparison of the social responsive-ness scale with the autism diagnostic interview-revised. Journal of Autism andDevelopmental Disorders 33 (4), 427–433.

Corbett, B.A., Carmean, V., Ravizza, S., Wendelken, C., Henry, M.L., Carter, C., Rivera,
S.M., 2009. A functional and structural study of emotion and face processing inchildren with autism. Psychiatry Research: Neuroimaging 173, 196–205.
Critchley, H.D., Daly, E.M., Bullmore, E.T., Williams, S.C.R., Van Amelsvoort, T., Robert-son, D.M., Rowe, A., Phillips, M., McAlonan, G., Howlin, P., Murphy, D.G.M., 2000.The functional neuroanatomy of social behaviour: changes in cerebral blood

9 Biobe

D

D

D

D

D

D

D

E

F

F

F

F

G

G

G

G

G

G

G

G

G

G

G

G

H

H


flow when people with autistic disorder process facial expressions. Brain 123,2203–2212.

alton, K.M., Nacewicz, B.M., Johnstone, T., Schaefer, H.S., Gernsbacher, M.A.,Goldsmith, H.H., Alexander, A.L., Davidson, R.J., 2005. Gaze fixation andthe neural circuitry of face processing in autism. Nature Neuroscience 8,519–526.

apretto, M., Davies, M.S., Pfeifer, J.H., Scott, A.A., Sigman, M., Bookheimer, S.Y.,Iacoboni, M., 2006. Understanding emotions in others: mirror neuron dysfunc-tion in children with autism spectrum disorders. Nature Neuroscience 9, 28–30.

eeley, Q., Daly, E.M., Surguladze, S., Page, L., Toal, F., Robertson, D., Curran, S.,Giampietro, V., Seal, M., Brammer, M.J., Andrew, C., Murphy, K., Phillips, M.L.,Murphy, D.G.M., 2007. An event related functional magnetic resonance imagingstudy of facial emotion processing in Asperger syndrome. Biological Psychiatry62, 207–217.

e Fossé, L., Hodge, S.M., Makris, N., Kennedy, D.N., Caviness, V.S., McGrath, L.,Steele, S., Ziegler, D.A., Herbert, M.R., Frazier, J.A., Tager-Flusberg, H., Harris, G.J.,2004. Language-association cortex asymmetry in autism and specific languageimpairment. Annals of Neurology 56, 757–766.

ichter, G.S., Belger, A., 2008. Atypical modulation of cognitive control by arousal inautism. Psychiatry Research: Neuroimaging 164, 185–197.

ichter, G.S., Belger, A., 2007. Social stimuli interfere with cognitive control inautism. Neuroimage 35, 1219–1230.

i Martino, A., Ross, K., Uddin, L.Q., Sklar, A.B., Castellanos, F.X., Milham, M.P., 2009.Functional brain correlates of social and nonsocial processes in autism spec-trum disorders: an activation likelihood estimation meta-analysis. BiologicalPsychiatry 65 (1), 63–74.

ickhoff, S.B., Laird, A.R., Grefkes, C., Wang, L.E., Zilles, K., Fox, P.T., 2009. Coordinate-based activation likelihood estimation meta-analysis of neuroimaging data: arandom-effects approach based on empirical estimates of spatial uncertainty.Human Brain Mapping 30 (9), 2907–2926.

loel, A., Ellger, T., Breitenstein, C., Knecht, S., 2003. Language perception activatesthe hand motor cortex: implications for motor theories of speech perception.The European Journal of Neuroscience 18 (3), 704–708.

ombonne, E., 2003. Epidemiological surveys of autism and other pervasive devel-opmental disorders: an update. Journal of Autism and Developmental Disorders33, 365–382.

ournier, K.A., Hass, C.J., Naik, S.K., Lodha, N., Cauruagh, J.H., 2010. Motor coordi-nation in autism spectrum disorders: a synthesis and meta-analysis. Journal ofAutism and Developmental Disorders 40 (10), 1227–1240.

reitag, C.M., Konrad, C., Häberlen, M., Kleser, C., von Gontard, A., Reith, W., Troje, N.F.,Krick, C., 2008. Perception of biological motion in autism spectrum disorders.Neuropsychologia 46, 1480–1494.

affrey, M.S., Kleinhans, N.M., Haist, F., Akshoomoff, N., Campbell, A., Courchesne,E., Müller, R.-A., 2007. A typical participation of visual cortex during word pro-cessing in autism: an fMRI study of semantic decision. Neuropsychologia 45,1672–1684.

allese, V., Fadiga, L., Fogassi, L., Rizzolatti, G., 1996. Action recognition in the pre-motor cortex. Brain 119, 593–609.

authier, I., Skudlarski, P., Gore, J.C., Anderson, A.W., 2000. Expertise for cars andbirds recruits brain areas involved in face recognition. Nature Neuroscience 3,191–197.

ervais, H., Belin, P., Boddaert, N., Leboyer, M., Coez, A., Sfaello, I., Barthelemy, C.,Brunelle, F., Samson, Y., Zilbovicius, M., 2004. Abnormal cortical voice processingin autism. Nature Neuroscience 7, 801–802.

eschwind, D.H., Levitt, P., 2007. Autism spectrum disorders: developmental dis-connection syndromes. Current Opinion in Neurobiology 17, 103–111.

ilbert, S.J., Meuwese, J.D.I., Towgood, K.J., Frith, C.D., Burgess, P.W., 2009. Abnormalfunctional specialization within medial prefrontal cortex in high-functioningautism: a multi-voxel similarity analysis. Brain 132, 869–878.

ilbert, S.J., Bird, G., Brindley, R., Frith, C.D., Burgess, P.W., 2008. Atypical recruitmentof medial prefrontal cortex in autism spectrum disorders: An fMRI study of twoexecutive function tasks. Neuropsychologia 46, 2281–2291.

omot, M., Belmonte, M.K., Bullmore, E.T., Bernard, F.A., Baron-Cohen, S., 2008. Brainhyper-reactivity to auditory novel targets in children with high-functioningautism. Brain 131, 2479–2488.

omot, M., Bernard, F.A., Davis, M.H., Belmonte, M.K., Ashwin, C., Bullmore, E.T.,Baron-Cohen, S., 2006. Change detection in children with autism: an auditoryevent-related fMRI study. Neuroimage 29, 475–484.

relotti, D.J., Klin, A.J., Gauthier, I., Skudlarski, P., Cohen, D.J., Gore, J.C., Volkmar,F.R., Schultz, R.T., 2005. fMRI activation of the fusiform gyrus and amygdala tocartoon characters but not to faces in a boy with autism. Neuropsychologia 43,373–385.

reimel, E., Schulte-Ruether, M., Kircher, T., Kamp-Becker, I., Remschmidt, H., Fink,G.R., Herpetz-Dahlmann, B., Konrad, K., 2009. Neural mechanisms of empathyin adolescents with autism spectrum disorder and their fathers. Neuroimage 49(1), 1055–1065.

rèzes, J., Wicker, B., Berthoz, S., de Gelder, B., 2009. A failure to grasp the affectivemeaning of actions in autism spectrum disorder subjects. Neuropsychologia 47,1816–1825.

adjikhani, N., Joseph, R., Manoach, D., Naik, P., Snyder, J., Dominick, K., Hoge, R., Vanden Stock, J., Tager Flusberg, H., de Gelder, B., 2009. Body expressions of emotion
do not trigger fear contagion in autism spectrum disorder. Social Cognition andAffective Neuroscience 4, 70–78.
adjikhani, N., Joseph, R.M., Snyder, J., Tager-Flusberg, H., 2007. Abnormal activationof the social brain during face perception in autism. Human Brain Mapping 28,441–449.


Hadjikhani, N., Joseph, R.M., Snyder, J., Chabris, C.F., Clark, J., Steele, S., McGrath,L., Vangel, M., Aharon, I., Feczko, E., 2004a. Activation of the fusiform gyruswhen individuals with autism spectrum disorder view faces. Neuroimage 22,1141–1150.

Hadjikhani, N.C.A., Chabris, C.F., Joseph, R.M., Clark, J., McGrath, L., Aharon, I., Feczko,E., Tager-Flusberg, H., Harris, G.J., 2004b. Early visual cortex organization inautism: an fMRI study. Neuroreport 15, 267–270.

Haist, F., Adamo, M., Westerfield, M., Courchesne, E., Townsend, J., 2005. Thefunctional neuroanatomy of spatial attention in autism spectrum disorder.Developmental Neuropsychology 27 (3), 425–458.

Harris, G.J., Chabris, C.F., Clark, J., Urban, T., Aharon, I., Steele, S., McGrath, L., Con-douris, K., Tager-Flusberg, H., 2006. Brain activation during semantic processingin autism spectrum disorders via functional magnetic resonance imaging. Brainand Cognition 61, 54–68.

Hauk, O., Johnsrude, I., Pulvermueller, F., 2004. Somatotopic representation of actionwords in human motor and premotor cortex. Neuron 41 (2), 301–307.

Haxby, J.V., Hoffman, E.A., Gobbini, M.I., 2002. Human neural systems for face recog-nition and social communication. Biological Psychiatry 51, 59–67.

Herbert, M.R., Ziegler, D.A., Deutsch, C.K., O’Brien, L.M., Lange, N., Bakardjiev, A.,Hodgson, J., Adrien, K.T., Steele, S., Makris, N., Kennedy, D., Harris, G.J., Caviness,V.S., 2003. Dissociations of cerebral cortex, subcortical and cerebral white mattervolumes in autistic boys. Brain 126, 1182–1192.

Herrington, J.D., Baron-Cohen, S., Wheelwright, S.J., Singh, K.D., Bullmore, E.T., Bram-mer, M., Williams, S.C.R., 2007. The role of MT+/V5 during biological motionperception in Asperger syndrome: an fMRI study. Research in Autism SpectrumDisorders 1, 14–27.

Hill, E.L., 2004. Executive dysfunction in autism. Trends in Cognitive Science 8 (1),26–32.

Holtmann, M., Boelte, S., Poustka, F., 2007. Autism spectrum disorders: sex dif-ferences in autistic domains and coexisting psychopathology. DevelopmentalMedicine and Child Neurology 49 (5), 361–366.

Hubl, D., Bolte, S., Feineis-Matthews, S., Lanfermann, H., Federspiel, A., Strik, W.,Poustka, F., Dierks, T., 2003. Functional imbalance of visual pathways indicatesalternative face processing strategies in autism. Neurology 61, 1232–1237.

Hughes, J.R., 2003. Autism: the first firm finding = underconnectivity? Epilepsy andBehaviour 11 (1), 20–24.

Humphreys, K., Hasson, U., Avidan, G., Minshew, N., Behrmann, M., 2008. Corticalpatterns of category-selective activation for faces, places and objects in adultswith autism. Autism Research 1, 52–63.

Itier, R.J., Batty, M., 2009. Neural bases of eye and gaze processing: the core of socialcognition. Neuroscience and Biobehavioural Reviews 33 (6), 843–863.

Just, M.A., Cherkassky, V.L., Keller, T.A., Kana, R.K., Minshew, N.J., 2007. Functionaland anatomical cortical underconnectivity in autism: evidence from an fMRIstudy of an executive function task and corpus callosum morphometry. CerebralCortex 17 (4), 951–961.

Just, M.A., Cherkassky, V.L., Keller, T.A., Kana, R.K., Minshew, N.J., 2006. Functionaland anatomical cortical underconnectivity in autism: evidence from an fMRIStudy of an executive function task and corpus callosum morphometry. CerebralCortex, bhl006.

Just, M.A., Cherkassky, V.L., Keller, T.A., Minshew, N.J., 2004. Cortical activation andsynchronization during sentence comprehension in high-functioning autism:evidence of underconnectivity. Brain 127, 1811–1821.

Kana, R.K., Keller, T.A., Cherkassky, V.L., Minshew, N.J., Just, M.A., 2006. Sentencecomprehension in autism: thinking in pictures with decreased functional con-nectivity. Brain 129, 2484–2493.

Kana, R.K., Keller, T.A., Cherkassky, V.L., Minshew, N.J., Just, M.A., 2009. Atypicalfrontal-posterior synchronization of Theory of Mind regions in autism duringmental state attribution. Social Neuroscience 4, 135–152.

Kana, R.K., Kellera, T.A., Minshew, N.J., Just, M.A., 2007. Inhibitory control in high-functioning autism: decreased activation and underconnectivity in inhibitionnetworks. Biological Psychiatry 62, 198–206.

Keehn, B., Brenner, L., Palmer, E., Lincoln, A.J., Muller, R.-A., 2008. Functional brainorganization for visual search in ASD. Journal of the International Neuropsycho-logical Society 14, 990–1003.

Kennedy, D.P., Courchesne, E., 2008. Functional abnormalities of the default net-work during self- and other-reflection in autism. Social Cognition and AffectiveNeuroscience 3, 177–190.

Kennedy, D.P., Redcay, E., Courchesne, E., 2006. Failing to deactivate: resting func-tional abnormalities in autism. Proceeds of the National Academy of Science USA103 (21), 8275–8280.

Kleinhans, N.M., Johnson, L.C., Richards, T., Mahurin, R., Greenson, J., Dawson, G.,Aylward, E., 2009. Reduced neural habituation in the amygdala and socialimpairments in autism spectrum disorders. American Journal of Psychiatry 166,467–475.

Kleinhans, N.M., Müller, R.-A., Cohen, D.N., Courchesne, E., 2008a. Atypical functionallateralization of language in autism spectrum disorders. Brain Research 1221,115–125.

Kleinhans, N.M., Richards, T., Sterling, L., Stegbauer, K.C., Mahurin, R., Johnson, L.C.,Greenson, J., Dawson, G., Aylward, E., 2008b. Abnormal functional connectivityin autism spectrum disorders during face processing. Brain 131, 1000–1012.

Kleinhans, N.M., Schweinsburg, B.C., Cohen, D.N., Müller, R.-A., Courchesne, E., 2007.
N-acetyl aspartate in autism spectrum disorders: regional effects and relation-ship to fMRI activation. Brain Research 1162, 85–97.
Knaus, T.A., Silver, A.M., Lindgren, K.A., Hadjikhani, N., Tager-Flusberg, H., 2008.fMRI activation during a language task in adolescents with ASD. Journal of theInternational Neuropsychological Society 14, 967–979.

Biobe

K

K

K

K

L

L

L

L

L

L

L

L

M

M

M

M

M

M

N

N

O

O

P

P

P

P

P


oshino, H., Kana, R.K., Keller, T.A., Cherkassky, V.L., Minshew, N.J., Just, M.A., 2008.fMRI investigation of working memory for faces in autism: visual coding andunderconnectivity with frontal areas. Cerebral Cortex 18, 289–300.

oshino, H., Carpenter, P.A., Minshew, N.J., Cherkassky, V.L., Keller, T.A., Just,M.A., 2005. Functional connectivity in an fMRI working memory task in high-functioning autism. Neuroimage 24, 810–821.

ringelbach, M.L., Rolls, E.T., 2004. The functional neuroanatomy of the humanorbitofrontal cortex: evidence from neuroimaging and neuropsychology.Progress in Neurobiology 72 (5), 341–372.

ringelbach, M.L., Rolls, E.T., 2003. Neural correlates of rapid reversal learning in asimple model of human social interaction. Neuroimage 20 (2), 1371–1383.

aird, A.R., Lancaster, J.L., Fox, P.T., 2005. BrainMap: the social evolution of a humanbrain mapping database. Neuroinformatics 3 (1), 65–78.

ee, P.S., Yerys, B.E., Della Rosa, A., Foss-Feig, J., Barnes, K.A., James, J.D., VanMeter,J., Vaidya, C.J., Gaillard, W.D., Kenworthy, L.E., 2009. Functional connectiv-ity of the inferior frontal cortex changes with age in children with autismspectrum disorders: a fcMRI study of response inhibition. Cerebral Cortex 19,1787–1794.

ee, J.E., Bigler, E.D., Alexander, A.L., Lazar, M., DuBray, M.B., Chung, M.K., Johnson, M.,Morgan, J., Miller, J.N., McMahon, W.M., Lu, J., Jeong, E.-K., Lainhart, J.E., 2007a.Diffusion tensor imaging of white matter in the superior temporal gyrus andtemporal stem in autism. Neuroscience Letters 424, 127–132.

ee, P.S., Foss-Feig, J., Henderson, J.G., Kenworthy, L.E., Gilotty, L., Gaillard, W.D.,Vaidya, C.J., 2007b. Atypical neural substrates of embedded figures task perfor-mance in children with autism spectrum disorder. Neuroimage 38, 184–193.

ord, C., Rutter, M., Goode, S., Heemsbergen, J., Jordan, H., Mawhood, L., Schopler,E., 1989. Autism diagnostic observation schedule: a standardized observationof communicative and social behavior. Journal of Autism and DevelopmentalDisorders 19 (2), 185–212.

ord, C., Rutter, M., Le Couteur, A., 1994. Autism Diagnostic Interview-Revised: arevised version of a diagnostic interview for caregivers of individuals with pos-sible pervasive developmental disorders. Journal of Autism and DevelopmentalDisorders 24 (5), 659–685.

oveland, K.A., Steinberg, J.L., Pearson, D.A., Mansour, R., Reddoch, S., 2008. Judg-ments of auditory-visual affective congruence in adolescents with and withoutautism: a pilot study of a new task using fMRI. Perceptual and Motor Skills 107,557–575.

una, B., Minshew, N.J., Garver, K.E., Lazar, N.A., Thulborn, K.R., Eddy, W.F., Sweeney,J.A., 2002. Neocortical system abnormalities in autism: an fMRI study of spatialworking memory. Neurology 59, 834–840.

anjaly, Z.M., Bruning, N., Neufang, S., Stephan, K.E., Brieber, S., Marshall, J.C., Kamp-Becker, I., Remschmidt, H., Herpertz-Dahlmann, B., Konrad, K., Fink, G.R., 2007.Neurophysiological correlates of relatively enhanced local visual search in autis-tic adolescents. Neuroimage 35, 283–291.

ason, R.A., Williams, D.L., Kana, R.K., Minshew, N., Just, M.A., 2008. Theory of Minddisruption and recruitment of the right hemisphere during narrative compre-hension in autism. Neuropsychologia 46, 269–280.

izuno, A., Villalobos, M.E., Davies, M.M., Dahl, B.C., Muller, R.-A., 2006. Partiallyenhanced thalamocortical functional connectivity in autism. Brain Research1104, 160–174.

uller, R.-A., Cauich, C., Rubio, M.A., Mizuno, A., Courchesne, E., 2004. Abnor-mal activity patterns in premotor cortex during sequence learning in autisticpatients. Biological Psychiatry 56, 323–332.

uller, R.-A., Kleinhans, N., Kemmotsu, N., Pierce, K., Courchesne, E., 2003. Abnor-mal variability and distribution of functional maps in autism: an fMRI study ofvisuomotor learning. American Journal of Psychiatry 160, 1847–1862.

uller, R.-A., Pierce, K., Ambrose, J.B., Allen, G., Courchesne, E., 2001. Atypical pat-terns of cerebral motor activation in autism: a functional magnetic resonancestudy. Biological Psychiatry 49, 665–676.

ewschaffer, C.J., Croen, L.A., Daniels, J., Giarelli, E., Grether, J.K., Levy, S.E., Mandell,D.S., Miller, L.A., Pinto-Martin, J., Reaven, J., Reynolds, A.M., Rice, C.E., Schen-del, D., Windham, G.C., 2007. The epidemiology of autism spectrum disorders.Annual Review of Public Health 28, 235–258.

oonan, S.K., Haist, F., Müller, R.-A., 2009. Aberrant functional connectivity inautism: evidence from low-frequency BOLD signal fluctuations. Brain Research1262, 48–63.

gai, M., Matsumoto, H., Suzuki, K., Ozawa, F., Fukuda, R., Uchiyama, I., Suckling, J.,Isoda, H., Mori, N., Takei, N., 2003. fMRI study of recognition of facial expressionsin high-functioning autistic patients. Neuroreport 14, 559–563.

ktem, F., Diren, B., Karaagaoglu, E., Anlar, B., 2000. Functional magnetic resonanceimaging in children with Asperger’s syndrome. Journal of Child Neurology 16,253–256.

elphrey, K.A., Morris, J.P., McCarthy, G., LaBar, K.S., 2007. Perception of dynamicchanges in facial affect and identity in autism. Social Cognition and AffectiveNeuroscience 2, 140–149.

elphrey, K.A., Morris, J.P., McCarthy, G., 2005. Neural basis of eye gaze processingdeficits in autism. Brain 128, 1038–1048.

ernet, C.R., Poline, J.B., Demonet, J.F., Rousselet, G.A., 2009. Brain classificationreveals the right cerebellum as the best biomarker of dyslexia. BMC Neuro-science 10, 67.

hilip, R.C.M., Whalley, H.C., Stanfield, A.C., Sprengelmeyer, R., Santos, I.M., Young,
A.W., Atkinson, A.P., Calder, A.J., Johnstone, E.C., Lawrie, S.M., Hall, J., 2010.Deficits in facial, body movement and vocal emotional processing in autismspectrum disorders. Psychological Medicine First View, 1–11.
ierce, K., Redcay, E., 2008. Fusiform function in children with an autism spectrumdisorder is a matter of “who”. Biological Psychiatry 64, 552–560.


Pierce, K., Haist, F., Sedaghat, F., Courchesne, E., 2004. The brain response to person-ally familiar faces in autism: findings of fusiform activity and beyond. Brain 127,2703–2716.

Pierce, K., Muller, R.A., Ambrose, J., Allen, G., Courchesne, E., 2001. Face processingoccurs outside the fusiform ‘face area’ in autism: evidence from functional MRI.Brain 124, 2059–2073.

Piggot, J., Kwon, H., Mobbs, D., Blasey, C., Lotspeich, L., Menon, V., Bookheimer,S., Reiss, A.L., 2004. Emotional attribution in high-functioning individuals withautistic spectrum disorder: a functional imaging study. Journal of the AmericanAcademy of Child and Adolescent Psychiatry 43, 473–480.

Pinkham, A.E., Hopfinger, J.B., Pelphrey, K.A., Piven, J., Penn, D.L., 2008. Neural basesfor impaired social cognition in schizophrenia and autism spectrum disorders.Schizophrenia Research 99, 164–175.

Probst, P., Leppert, T., 2008. Brief report: outcomes of a teacher training program forautism spectrum disorders. Journal of Autism and Developmental Disorders 38,1791–1796.

Rajendran, G., Mitchell, P., 2007. Cognitive theories of autism. DevelopmentalReview 27, 224–260.

Redcay, E., Courchesne, E., 2008. Deviant functional magnetic resonance imagingpatterns of brain activity to speech in 2–3-year-old children with autism spec-trum disorder. Biological Psychiatry 64, 589–598.

Redcay, E., Courchesne, E., 2005. When is the brain enlarged in autism? A meta-analysis of all brain size reports. Biological Psychiatry 58 (1), 1–9.

Ring, H.A., Baron-Cohen, S., Wheelwright, S., Williams, S.C.R., Brammer, M., Andrew,C., Bullmore, E.T., 1999. Cerebral correlates of preserved cognitive skills inautism: a functional MRI study of embedded figures task performance. Brain122, 1305–1315.

Sahyoun, C., Soulières, I., Belliveau, J., Mottron, L., Mody, M., 2009. Cognitive differ-ences in pictorial reasoning between high-functioning autism and Asperger’ssyndrome. Journal of Autism and Developmental Disorders 39, 1014–1023.

Samson, F., Mottron, L., Soulieres, I., Zeffiro, T.A., 2011. Enhanced visual functioningin autism: an ALE meta-analysis. Human Brain Mapping [Epub ahead of print].

Schmitz, N., Rubia, K., van Amelsvoort, T., Daly, E., Smith, A., Murphy, D.G.M., 2008.Neural correlates of reward in autism. British Journal of Psychiatry 192, 19–24.

Schmitz, N., Rubia, K., Daly, E., Smith, A., Williams, S., Murphy, D.G.M., 2006. Neu-ral correlates of executive function in autistic spectrum disorders. BiologicalPsychiatry 59, 7–16.

Schopler, E., Reichler, R.J., DeVellis, R.F., Daly, K., 1980. Toward objective classifi-cation of childhood autism: Childhood Autism Rating Scale (CARS). Journal ofAutism and Developmental Disorders 10 (1), 91–103.

Schulte-Rüther, M., Markowitsch, H.J., Fink, G.R., Piefke, M., 2007. Mirror neuronand Theory of Mind mechanisms involved in face-to-face interactions: a func-tional magnetic resonance imaging approach to empathy. Journal of CognitiveNeuroscience 19, 1354–1372.

Schultz, R.T., Gauthier, I., Klin, A., Fulbright, R.K., Anderson, A.W., Volkmar, F., Skud-larski, P., Lacadie, C., Cohen, D.J., Gore, J.C., 2000. Abnormal ventral temporalcortical activity during face discrimination among individuals with autism andAsperger syndrome. Archives of General Psychiatry 57, 331–340.

Shafritz, K.M., Dichter, G.S., Baranek, G.T., Belger, A., 2008. The neural circuitrymediating shifts in behavioral response and cognitive set in autism. BiologicalPsychiatry 63, 974–980.

Silani, G., Bird, G., Brindley, R., Singer, T., Frith, C., Frith, U., 2008. Levels ofemotional awareness and autism: an fMRI study. Social Neuroscience 3,97–112.

Silk, T.J., Rinehart, N., Bradshaw, J.L., Tonge, B., Egan, G., O’Boyle, M.W., Cunning-ton, R., 2006. Visuospatial processing and the function of prefrontal-parietalnetworks in autism spectrum disorders: a functional MRI study. The AmericanJournal of Psychiatry 163, 1440–1443.

Soulieres, I., Zeffiro, T.A., Girard, M.L., Mottron, L., 2011. Enhanced mental imagemapping in autism. Neuropsychologia 49 (5), 848–857.

Sparrow, S.S., Cicchetti, D.V., 1985. Diagnostic uses of the Vineland Adaptive Behav-ior Scales. Journal of Pediatric Psychology 10 (2), 215–225.

Stanfield, A.C., McIntosh, A.M., Spencer, M.D., Philip, R., Gaur, S., Lawrie, S.M., 2008.Towards a neuroanatomy of autism: a systematic review and meta-analysisof structural magnetic resonance imaging studies. European Psychiatry 23,289–299.

Steffenburg, S., Steffenburg, U., Gillberg, C., 2003. Autism spectrum disorders inchildren with active epilepsy and learning disability: comorbidity, pre- and peri-natal background, and seizure characterics. Developmental Medicine and Childneurology 45 (11), 724–730.

Stoodley, C.J., 2011. The cerebellum and cognition: evidence from functional imagingstudies. Cerebellum [Epub ahead of print].

Takarae, Y., Minshew, N.J., Luna, B., Sweeney, J.A., 2007. Atypical involvement of fron-tostriatal systems during sensorimotor control in autism. Psychiatry Research:Neuroimaging 156, 117–127.

Takeuchi, M., Harada, M., Matsuzaki, K., Nishitani, H., Mori, K., 2004. Difference ofsignal change by a language task on autistic patients using functional MRI. TheJournal of Medical Investigation 51, 59–62.

Thakkar, K.N., Polli, F.E., Joseph, R.M., Tuch, D.S., Hadjikhani, N., Barton, J.J.S.,Manoach, D.S., 2008. Response monitoring, repetitive behaviour and anteriorcingulate abnormalities in ASD. Brain, awn099.

Turkeltaub, P.E., Flowers, D.L., Verbalis, A., Miranda, M., Gareau, L., Eden, G.F., 2004.The neural basis of hyperlexic reading: an fMRI case study. Neuron 41, 11–25.

Turner, K., Frost, L., Linsenbardt, D., McIlroy, J., Muller, R.-A., 2006. Atypically diffusefunctional connectivity between caudate nuclei and cerebral cortex in autism.Behavioural and Brain Functions 2, 34.

9 Biobe

U

V

W

W

W

W


ddin, L.Q., Davies, M.S., Scott, A.A., Zaidel, E., Bookheimer, S.Y., Iacoboni, M.,Dapretto, M., 2008. Neural basis of self and other representation in autism: anfMRI study of self-face recognition. PLoS ONE 3, e3526.

illalobos, M.E., Mizuno, A., Dahl, B.C., Kemmotsu, N., Muller, R.-A., 2005.Reduced functional connectivity between V1 and inferior frontal cor-tex associated with visuomotor performance in autism. Neuroimage 25,916–925.

ang, L., Liu, X., Guise, K.G., Knight, R.T., Ghajar, J., Fan, J., 2010. Effective connectivityof the fronto-parietal network during attentional control. Journal of CognitiveNeuroscience 22 (3), 543–553.

ang, A.T., Lee, S.S., Sigman, M., Dapretto, M., 2007. Reading affect in the face andvoice: neural correlates of interpreting communicative intent in children andadolescents with autism spectrum disorders. Archives of General Psychiatry 64,698–708.

ang, A.T., Lee, S.S., Sigman, M., Dapretto, M., 2006. Neural basis of irony compre-hension in children with autism: the role of prosody and context. Brain 129,932–943.

ang, A.T., Dapretto, M., Hariri, A.R., Sigman, M., Bookheimer, S.Y., 2004. Neuralcorrelates of facial affect processing in children and adolescents with autism


spectrum disorder. Journal of the American Academy of Child and AdolescentPsychiatry 43, 481–490.

Welchew, D.E., Ashwin, C., Berkouk, K., Salvador, R., Suckling, J., Baron-Cohen, S.,Bullmore, E., 2005. Functional disconnectivity of the medial temporal lobe inAsperger’s syndrome. Biological Psychiatry 57, 991–998.

Wicker, B., Fonlupt, P., Hubert, B., Tardif, C., Gepner, B., Deruelle, C., 2008. Abnormalcerebral effective connectivity during explicit emotional processing in adultswith autism spectrum disorder. Social Cognition and Affective Neuroscience 3,135–143.

Williams, J.H.G., Waiter, G.D., Gilchrist, A., Perrett, D.I., Murray, A.D., Whiten, A.,2006. Neural mechanisms of imitation and ‘mirror neuron’ functioning in autis-tic spectrum disorder. Neuropsychologia 44, 610–621.

Woodbury-Smith, M.R., Robinson, J., Wheelwright, S., Baron-Cohen, S., 2005. Screen-ing adults for Asperger Syndrome using the AQ: a preliminary study of its
diagnostic validity in clinical practice. Journal of Autism and DevelopmentalDisorders 35 (3), 331–335.
Zilbovicius, M., Meresse, I., Chabane, N., Brunelle, F., Samson, Y., Boddaert, N., 2006.Autism, the superior temporal sulcus and social perception. Trends in Neuro-sciences 29 (7), 359–366.

neuroscience and biobehavioral reviewsbrainmap.org/pubs/philip_nbr_12.pdf · 2015. 1. 6. · r.c.m....

Documents